Ba elheime , Ch is ian e al.
A icle — Published Ve sion
Concep ualizing hyb id in elligen se ice ecosys ems
Elec onic Ma ke s
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Ba elheime , Ch is ian e al. (2025) : Concep ualizing hyb id in elligen se ice
ecosys ems, Elec onic Ma ke s, ISSN 1422-8890, Sp inge , Be lin, Heidelbe g, Vol. 35, Iss. 1,
h ps://doi.o g/10.1007/s12525-025-00798-4
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/323687
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Vol.:(0123456789)
Elec onic Ma ke s (2025) 35:63
h ps://doi.o g/10.1007/s12525-025-00798-4
RESEARCH PAPER
Concep ualizing hyb id in elligen se ice ecosys ems
Ch is ianBa elheime 1 · DanielHeinz2· Sa ahHönigsbe g3· DominikSiemon4· MaheiManhaiLi5·
TimoS ohmann6· JensPoeppelbuss7· Ch is ophPe e s8
Recei ed: 29 May 2024 / Accep ed: 16 May 2025
© The Au ho (s) 2025
Abs ac
Wi h he p oli e a ion o a i icial in elligence (AI) echnologies, he collabo a ion o human and AI ac o s in alue co-
c ea ion p ocesses pe mea es a ious applica ion domains. In his concep ual pape , we in eg a e concep s om human-AI
collabo a ion and se ice esea ch and p esen a concep ual amewo k o hyb id in elligen se ice ecosys ems (HISE).
The amewo k ex ends he exis ing concep ualiza ions o se ice ecosys ems as pu o wa d by he se ice-dominan logic
(S-D logic) by emphasizing how ac o s delibe a ely con igu e human and a i icial agencies o co-c ea e alue ia hyb id
in elligen se ice exchange and how his impac s ecosys em o ma ion and e olu ion. Ou concep ualiza ion highligh s ha
alue co-c ea ion in HISE is guided and acili a ed by sha ed esou ces and ins i u ional a angemen s, which di e om
p e ious se ice ecosys ems h ough he eme gence o hyb id agency. We demons a e he applicabili y o ou amewo k
wi h i e illus a i e HISE scena ios and p o ide i e heo e ical p oposi ions. Ou indings ex end exis ing knowledge by
heo izing on how o inco po a e hyb id in elligence in o alue co-c ea ion p ocesses. The eby, we p o ide a ounda ion o
u u e in e disciplina y esea ch on human-AI collabo a ion a he in e sec ion o in o ma ion sys ems, human–compu e
in e ac ion, and se ice esea ch wi h S-D logic as a uni ying heo e ical lens.
Keywo ds Hyb id in elligence· Se ice ecosys ems· Value co-c ea ion· A i icial in elligence
JEL Classi ica ion M15
In oduc ion
Wi h he p oli e a ion o a i icial in elligence (AI) echnolo-
gies, we obse e a undamen al shi in how people in e ac
wi h hese echnologies indi idually and collec i ely in se -
ice ecosys ems (Huang & Rus , 2018). This e olu ion ena-
bles le e aging he po en ial o hyb id in elligence by allo-
ca ing asks o ei he human o AI agen s, depending on hei
espec i e s eng hs (Delle mann e al., 2019). The wide-
sp ead adop ion o gene a i e AI applica ions, such as Cha -
GPT, highligh s hyb id in elligence’s ans o ma i e impac
on alue co-c ea ion p ocesses. By s a egically assigning
asks based on he con ex , hyb id in elligence enables ou -
comes ha nei he humans no AI could achie e indepen-
den ly (Hemme e al., 2023). In pa icula , he abili y o
employ “AI as a se ice” is expec ed o con inuously uel
he g ow h o hyb id in elligen se ice oppo uni ies (New-
lands, 2021) because i allows indi iduals o access and use
AI capabili ies wi hou he need o signi ican in es men in
in as uc u e and expe ise. Howe e , despi e ad ancemen s
Responsible Edi o : F ancesco Polese.
* Ch is ian Ba elheime
[email p o ec ed]
1 Uni e si y o Gö ingen, Pla z de Gö inge Sieben 5,
37073Gö ingen, Ge many
2 Ka ls uhe Ins i u e o Technology, Kaise s aße 12,
76131Ka ls uhe, Ge many
3 ICN Business School, Uni e si y o Lo aine, CEREFIGE,
F-54000Nancy, F ance
4 LUT Uni e si y, Yliopis onka u 34, 53850Lappeen an a,
Finland
5 Uni e si y o S . Gallen (HSG), Mülle -F iedbe g-S asse 8,
9000S .Gallen, Swi ze land
6 Uni e si y o Müns e , Leona do-Campus348149Müns e ,
Ge many
7 Ruh Uni e si y Bochum, Uni e si ä ss aße 150,
44801Bochum, Ge many
8 Uni e si y o heBundesweh Munich,
We ne -Heisenbe g-Weg 39, 85577Neubibe g, Ge many
Elec onic Ma ke s (2025) 35:63 63 Page 2 o 27
in esol ing adi ional echnical cons ain s ela ed o AI
deploymen (e.g., scalabili y, cos ) and egula o y cons ain s
conce ning AI sa e y (e.g., go e nance o high- isk AI sys-
ems), he ull po en ial o human-AI collabo a ion in co-
c ea ing alue emains unde u ilized. This dispa i y sugges s
he need o a amewo k ha helps desc ibe and unde s and
how humans and AI co-c ea e alue in se ice ecosys ems
h ough in e ac ion and esou ce in eg a ion.
The se ice-dominan logic o ma ke ing (S-D logic),
which is based on he undamen al assump ion ha se ice is
he basis o all economic exchange, p o ides an es ablished
heo e ical lens o s udy alue co-c ea ion h ough se ice
exchange be ween ac o s in se ice ecosys ems (Va go &
Lusch, 2004, 2016). A se ice ecosys em is a “ ela i ely
sel -con ained, sel -adjus ing sys em o esou ce-in eg a ing
ac o s connec ed by sha ed ins i u ional a angemen s and
mu ual alue c ea ion h ough se ice exchange” (Va go &
Lusch, 2016, pp. 10–11). As such, i ep esen s “a spon ane-
ously sensing and esponding spa ial and empo al s uc u e
o la gely loosely coupled, alue-p oposing social and eco-
nomic ac o s in e ac ing h ough ins i u ions, echnology,
and language o (1) co-p oduce se ice o e ings, (2) engage
in mu ual se ice p o ision, and (3) co-c ea e alue” (Va go
& Lusch, 2011, p. 185). This iew o se ice ecosys ems
emphasizes he impo ance o ins i u ions in alue co-c e-
a ion and se ice inno a ion. I explica es he complex and
dynamic na u e o he social sys ems h ough which se ice
is p o ided (Va go & Akaka, 2012). Fu he mo e, i assumes
ha “se ice ecosys ems a e sys ems o sys ems in which he
a ious sys ems in e ac , and di e en le els o analysis can
be applied: mic o (ac o engagemen ), meso (se s o ac o s
and hei esou ces), and mac o (ecosys em and ins i u ional
logics)” (S o backa e al., 2016, p. 3009). Va go and Lusch
(2011) u he emphasize ha echnology de elopmen , in
pa icula , d i es he e olu ion and pe o mance o se ice
ecosys ems, enabling ac o s o sense and espond mo e and
mo e spon aneously.
The ascendance o AI echnologies and he g owing
impo ance o human-AI collabo a ion in bo h day- o-day
wo k p ac ices and p i a e con ex s a e o ces ha shape
se ice ecosys em e olu ion. Fo example, he apid de el-
opmen o gene a i e AI echnologies has led o a su ge o
AI inno a ions (e.g., cha bo s, image gene a ion) en e ing
he ma ke , impac ing he ins i u ional logics (a he mac o
le el). This in lux o new echnologies upda es he esou ce
s uc u es o ac o s by in eg a ing mo e ad anced AI ools
in o he se ice ecosys em, he eby al e ing esou ce in e-
g a ion pa e ns o ac o s (a he meso le el), and c ea es
hyb id ac o s who alloca e asks o human o AI agencies (a
he mic o le el). In esponse o hese de elopmen s, egu-
la o y bodies like he Eu opean Union a e p oposing and
implemen ing legisla ion such as he EU AI Ac (Eu opean
Union, 2024) o egula e AI use in ended o eshape exis ing
ins i u ional a angemen s (a he mac o le el), ensu ing e h-
ical s anda ds and sa e y. Consequen ly, o ganiza ions like
Me a decided o wi hd aw ce ain AI models and p oduc s
om he EU ma ke , which limi s he esou ces a ailable o
EU ci izens and po en ially inhibi s se ice exchange and
alue co-c ea ion (a he meso le el). This example illus-
a es how ins i u ional a angemen s on he mac o le el bo h
guide meso-le el esou ce in eg a ion pa e ns and e ol e
h ough mic o-le el ac o in e ac ions, con inuously shaping
and eshaping he dynamics o alue co-c ea ion in hyb id
in elligen se ice ecosys ems (HISE).
While many human-AI collabo a ions a e inhe en ly
se ice-o ien ed (Kno e e al., 2021), and he e a e al eady
s udies ha ha e explo ed AI’s ole in speci ic se ice con-
ex s such as ma ke ing (e.g., Wedel & Kannan, 2016), cus-
ome se ice (e.g., Adam e al., 2021; Kno e e al., 2021),
and b oade se ice esea ch (e.g., Wi z e al., 2018), he e
emains a signi ican lack o unde s anding o how human
and AI agencies can be con igu ed o op imize esou ce
in eg a ion and alue co-c ea ion wi hin se ice ecosys-
ems (B eidbach & B odie, 2017). This unde - esea ched
a ea limi s ou abili y o s udy and design cu en and u u e
se ice ecosys ems e ec i ely and o p edic hei e olu ion.
Add essing his issue is c i ical and equi es a concep ual
amewo k o na iga e he e ol ing esou ces and shi -
ing ins i u ional a angemen s in hese se ice ecosys ems,
which al e esou ce in eg a ion pa e ns in se ice exchange
o e ime.
Despi e signi ican con ibu ions om bo h se ice
esea ch (e.g., Kno e e al., 2021; Wi z e al., 2018) and
hyb id in elligence esea ch (e.g., Delle mann e al., 2019),
he e is a lack o in eg a ion be ween hese ields ha lea es
he phenomenon o HISE unde -concep ualized. To build
human-cen e ed solu ions aligned wi h mode n impe a i es
o alue co-c ea ion and collabo a ion, we need in eg a ed
knowledge ha le e ages he s eng hs o bo h domains o
unde s and how o sys ema ically desc ibe, explain, analyze,
and design human-AI collabo a ion in se ice ecosys ems.
Un esol ed ques ions abou HISE e e o he a ious se -
ice ecosys em elemen s and hei in e play, including, e.g.,
how ac o s and hei esou ces a e con igu ed and e ol e in
HISE, how o design and manage se ice ecosys ems ha
enable hyb id in elligen se ice exchange, and wha ole
ins i u ional a angemen s play in his ega d.
The lack o a comp ehensi e unde s anding o HISE is
no jus ele an om a heo e ical pe spec i e, bu i can
also ha e p o ound implica ions ac oss di e se applica ion
domains (e.g., mobili y, elde ly ca e, so wa e enginee -
ing, ag icul u e, IT cus ome suppo ). This knowledge gap
can lead o a nega i e bias owa d human-AI collabo a ion
despi e i s po en ial o os e ing mo e e ec i e and sus-
ainable u u e se ice ecosys ems (Dwi edi e al., 2023).
Fo example, he e is an e osion o us due o unclea
Elec onic Ma ke s (2025) 35:63 Page 3 o 27 63
cos -bene i analyses o hyb id in elligen applica ions and
a lack o anspa ency conce ning hei employmen (Gkinko
& Elbanna, 2023; Hildeb and & Be gne , 2021; Schue z
& Venka esh, 2020). Con e sely, he epo ed signi ican
g ow h o AI applica ions in sec o s like e ail, heal hca e,
IT and elecommunica ions, manu ac u ing, ene gy (Mo do
In elligence, 2024), and public se ices (He nandez, 2022)
highligh s he po en ial bene i s o ganiza ions pe cei e in
in eg a ing AI in o hei alue co-c ea ion p ocesses. Hence,
unde s anding and concep ualizing alue co-c ea ion in such
se ings is c i ical o esea che s and p ac i ione s aiming o
design and manage se ice ecosys ems e ec i ely.
The e o e, his pape p oposes a concep ual amewo k
o ad ance ou unde s anding o HISE. By adop ing a se -
ice ecosys em pe spec i e ha is oo ed in S-D logic, we
aim o desc ibe he complex in e play be ween di e en
kinds o ac o s—including humans, AI, and hyb ids—and
hei impac on alue co-c ea ion p ac ices and ins i u ional
a angemen s in HISE. The eby, we explo e he in e ela-
ionships o alue co-c ea ion, esou ces, and hei mobi-
liza ion, as well as ins i u ional a angemen s conce ning
AI-d i en se ice o e ings and se ice in e ac ions. In
his pape , we de ine HISE as se ice ecosys ems ha le -
e age human and AI o con igu e human and AI agencies
as human-AI hyb ids. By employing an S-D logic lens, we
unde s and HISE as e ol ing sys ems ha con inuously
e ol e om he con igu a ions o agency h ough ac o s by
os e ing he e olu ion o bo h human and a i icial agen-
cies, imp o ing he e iciency and e ec i eness o hyb id
in elligen se ice. In his iew, HISE a e able o con inu-
ously imp o e he co-c ea ed alue o all ac o s as po en ial
bene icia ies.
By concep ualizing HISE, we make h ee key con i-
bu ions o he academic knowledge base: Fi s , he HISE
amewo k is he i s o p o ide a concep ual ounda ion o
s udy phenomena a he in e sec ion o human-AI collabo a-
ion and alue co-c ea ion in se ice ecosys ems. Second, i
equips esea che s and manage s wi h a ool o unde s and,
design, and manage HISE as s a e-o - he-a se ice eco-
sys ems ha le e age he po en ial o AI. Thi d, we p esen
i e p oposi ions alongside he concep ualiza ion ha guide
high-impac u u e esea ch on HISE, building on ou p o-
posed amewo k.
We demons a e he e sa ili y and applicabili y o ou
amewo k h ough a a ie y o scena ios—semi-au ono-
mous d i ing, elde ly ca e, sus ainable coding, p ecision
ag icul u e, ITcus ome suppo — ha encompass di e se
se ice ecosys ems ha s em om a wide ange o indus-
ies, om sma p oduc s like semi-au onomous ehicles
o human-cen e ed se ices such as elde ly ca e, and om
in angible se ices like so wa e de elopmen and IT sup-
po o angible a ming p ocesses in ag icul u e. These sce-
na ios cap u e a b oad spec um o se ice in e ac ions and
ecosys em dynamics, showcasing he amewo k’s lexibili y
o be applied o di e se hyb id se ice encoun e s in ol ing
ac o s such as echnology p o ide s, se ice p o ide s, end
use s, egula o s, and en i onmen al ac o s. While p e ious
s udies ha e ocused on AI’s ans o ma i e ole in speci ic
sec o s like ou ism (So aya González-Mendes e al., 2024),
ou concep ual s udy b oadens his pe spec i e by explo ing
he ole o hyb id in elligence ac oss a ious indus ies.
The emainde o he pape is s uc u ed as ollows:
Sec .2 desc ibes he heo e ical unde pinnings o he s udy,
ou lining human-AI collabo a ion, hyb id in elligence, se -
ice ecosys ems om a S-D logic pe spec i e, and agency
esea ch in in o ma ion sys ems (IS). In Sec .3, we ou line
ou beha io al concep ual esea ch app oach, which s uc-
u es he subsequen Sec .4, in which we concep ualize
HISE and apply he amewo k o i e illus a i e scena ios.
In Sec .5, we p esen i e p oposi ions ha can guide u u e
esea ch on HISE be o e concluding he pape in Sec .6.
Theo e ical backg ound
Hyb id in elligence andhuman‑AI collabo a ion
Hyb id in elligence e e s o he combina ion o human
and AI agen s, le e aging hei complemen a y s eng hs o
o m a socio- echnical ensemble (Delle mann e al., 2019;
Malone, 2018). I is he ou come o human-AI collabo a ion
(Fügene e al., 2022). This iew on human-AI collabo a ion
ecognizes ha AI sys ems ha e unique capabili ies, such as
p ocessing la ge amoun s o da a, pa e n ecogni ion, na u-
al language p ocessing, image ecogni ion, and p edic i e
analysis (Good ellow e al., 2016; LeCun e al., 2015; Rus-
sell & No ig, 2021), which can complemen human in elli-
gence in a ious asks (Da enpo & Ki by, 2016). Simila ly,
human in elligence p o ides c ea i i y, empa hy, and con ex-
ual unde s anding, which can complemen he limi a ions o
AI sys ems (B ynjol sson & McA ee, 2014), such as biases
a ising om aining da a (Caliskan e al., 2017) o lack o
explainabili y in hei decision-making p ocesses (A ie a
e al., 2019). Howe e , hyb id in elligence does no simply
in ol e inse ing human in elligence in o he AI loop o o
au oma e simple asks h ough machine lea ning. Ra he ,
i seeks o sol e complex p oblems by delibe a ely alloca -
ing and coo dina ing asks among he e ogeneous algo i h-
mic and human agen s, he eby enabling ac ual human-AI
collabo a ion.
Hyb id in elligence esea ch has sough o explo e he
in e play be ween human agen s and AI agen s in a ious
con ex s, such as decision-making (Ja ahi, 2018), p ob-
lem-sol ing (Woolley e al., 2010), and eam pe o mance
(An hony e al., 2023; Siemon e al., 2022). Howe e , e en
capabili ies p e iously associa ed wi h humans, such as
Elec onic Ma ke s (2025) 35:63 63 Page 4 o 27
c ea i e ac i i ies, can be enhanced in a hyb id in elligence
se ing o AI sys ems oge he wi h humans (Jia e al., 2023;
Siemon e al., 2022). A key insigh om his esea ch s eam
is ha he collabo a ion be ween human and AI agen s can
lead o ou comes ha a e supe io o wha ei he agen
could achie e indi idually (Delle mann e al., 2019; Hem-
me e al., 2023). Fo ins ance, Woolley e al. (2010) ound
ha hyb id eams can ou pe o m eams o human agen s
wo king alone on complex p oblem-sol ing asks. In addi-
ion, S iegli z e al. (2021), o example, s udy he e ec s
o human beha io in e ms o social loa ing o delega ing
esponsibili y o an AI in a eam. Siemon e al. (2022) show
how humans in a collabo a i e c ea i e p ocess eac o he
c i ical oices o an AI on hei own ideas.
Simila ly, Fügene e al. (2022) co obo a e h ough hei
expe imen al s udy ha human-AI ensembles pe o m supe-
io ly in classi ica ion asks when compa ed o he esul s
achie ed by bo h ac o s indi idually. O pa icula ele ance,
howe e , is ha his ensemble only achie es supe io pe -
o mance when he AI can ac i ely delega e asks o he
human and hus has i s own agency o in luence he collabo-
a ion dynamics. Fu he mo e, Gkinko and Elbanna (2023)
show ha sus ainable hyb id in elligen wo k scena ios only
eme ge when AI is pe cei ed as a pe sonal assis an wi h
i s own capabili ies and scope o ac ion and no jus as a
ool. These ecen s udies indica e ha AI agency plays an
inc easingly ac i e and equi able ole in hyb id in elligence
se ings.
Addi ionally, ea ly pionee ing s udies exis ha desc ibe
he e ec s o AI in ecosys ems, o example, conce n-
ing e hical implica ions o AI o inno a ion ecosys ems
(S ahl, 2022), da a ne wo k e ec s, use decen aliza ion
on da a- and AI-d i en digi al pla o ms (Clough & Wu,
2022; G ego y e al., 2022). Howe e , hese s udies do no
add ess hyb id in elligence o human-AI collabo a ion spe-
ci ically. Mo eo e , while esea ch on hyb id in elligence has
p o ided aluable insigh s in o he po en ial o human-AI
collabo a ion, i has p ima ily ocused on he indi idual and
eam le els (including hei dynamics and in luences) (e.g.,
Delle mann e al., 2019). Fo example, Recke e al. (2023)
p opose he size and he e ogenei y o a human-AI ensemble
(one- o-one, one- o-many, many- o-one, o many- o-many),
i s con ol sequence (human- i s , machine- i s , o synch o-
nous), he na u e o he p oblem domain (well-s uc u ed
o poo ly s uc u ed), and he o e a ching inno a ion goal
(p oblem-d i en o solu ion-d i en) as impo an aspec s
ha de e mine ask alloca ion in he agency con igu a ion
p ocess. Fab i e al. (2023) p esen a axonomy o human-AI
hyb ids, which we would equa e wi h he con igu ed agency
in ou amewo k, and iden i y i e con igu a ion a che-
ypes (AI p e-wo ke , ou sou cing AI, supe powe -gi ing
AI, assembly-line AI, collabo a o AI) based on clus e ing
101 human-AI hyb ids. Howe e , he agency con igu a ion
o human and AI agencies is s ill a nascen esea ch opic
ha is adically changing due o new echnological ad ance-
men s, new o ms o deep in eg a ion o AI in o p e-exis ing
p ac ices, and he ongoing ans o ma ion o hese p ac ices
and alue co-c ea ion pa e ns in gene al.
To conclude, ega ding ou aim o concep ualizing HISE,
esea ch o da e in his ield has p ima ily ocused on mic o-
le el in e ac ions be ween humans and AI bu has no su -
icien ly conside ed he in e play wi h he meso and mac o
le els o se ice ecosys ems in which humans and AI in e -
ac . In pa icula , he hyb id in elligence li e a u e has no
ully add essed how he collabo a ion be ween human and
AI agen s can in luence alue co-c ea ion, esou ces, and
ins i u ional a angemen s, which a e cen al elemen s in
se ice ecosys ems.
Se ice‑dominan logic andse ice ecosys ems
S-D logic eme ged as a ounda ional pa adigm shi in ma -
ke ing and se ice esea ch, o e ing a well-accep ed heo e -
ical lens wi h which o s udy economic exchange (Va go &
Lusch, 2004). I posi ions all economic exchange as se ice-
o -se ice o skill- o -skill exchange among ac o s, wi h
emphasis on b oade socie al and economic sys ems (Va go
& Lusch, 2016). In his ega d, Lusch and Va go (2006, p.
283) ( e-)de ine se ice as “ he applica ion o specialized
compe encies (knowledge and skills), h ough deeds, p o-
cesses, and pe o mances o he bene i o ano he en i y
o he en i y i sel .”
Since i s incep ion, S-D logic has unde gone u he
de elopmen , leading o e isions o i s ounda ional p em-
ises and hei consolida ion in o i e axioms. As pa o an
“ins i u ional and dyad- o-ne wo k- o-sys ems u n” (Va go
& Lusch, 2016, p. 6), ecen wo k highligh s he impo -
ance o a se ice ecosys em iew “ o allow a mo e holis ic,
dynamic, and ealis ic pe spec i e o alue c ea ion, h ough
exchange, among a wide , mo e comp ehensi e ( han i m
and cus ome ) con igu a ion o ac o s” (Va go & Lusch,
2016, p. 5 .)
Acco dingly, S-D logic ea s all pa icipan s in economic
and social exchange as gene ic ac o s (Ekman e al., 2016;
Hönigsbe g & Din e , 2024; Wieland e al., 2012) who adop
a ious oles in alue co-c ea ion p ocesses a he han being
elega ed o adi ional ca ego ies such as “supplie ” o “cus-
ome ” (Va go & Lusch, 2016, 2017). An ac o can hus be
any ma ke pa icipan in ol ed in ac o - o-ac o exchanges
(Va go & Lusch, 2016). Hence, ac o s can be indi iduals,
o ganiza ions, o g oups (S o backa e al., 2016; Wieland
e al., 2012) and can be nes ed in collec ions o ac o s, like a
g oup o indi iduals whe e bo h he g oup and i s indi idu-
als a e concep ualized as ac o s om an S-D logic pe spec-
i e (Va go & Lusch, 2011). Ac o s can be en i ies wi hin
(e.g., a depa men , unc ion, o local b anches) o ex e nal
Elec onic Ma ke s (2025) 35:63 Page 5 o 27 63
o an o ganiza ion (e.g., ano he company) (Schymanie z &
Jonas, 2020). S o backa e al. (2016) u he sugges ha ,
wi h ad ancing echnology, machines can also be conside ed
ac o s, ex ending he scope o ac o - o-ac o in e ac ions.
Recognizing his mul iplici y o in e wined co-c ea ing
ac o s—indi iduals, o ganiza ions, o o he se ice sys-
ems—shi s he ocus om alue c ea ion as a linea , i m-
cen ic p ocess o a mo e dynamic and in e ac i e p ocess
in ol ing a ious ac o s wi hin a se ice ecosys em (Chan-
dle & Lusch, 2015). Fo ins ance, en i ies such as ci ies,
indus ies, and ma ke s can hemsel es be iewed as se ice
ecosys ems (Sa no e al., 2024).
In he se ice science li e a u e, ac o s a e commonly
concep ualized as “se ice sys ems” (Va go & Lusch, 2011,
p. 186). This concep encompasses en i ies a any le el o
agg ega ion— om single indi iduals as a omic se ice sys-
ems o en i e communi ies o o ganiza ions (Spoh e e al.,
2008; Va go & Lusch, 2011). While i may be coun e in ui-
i e o label a single indi idual a “sys em,” om a se ice
science pe spec i e, each pe son in eg a es esou ces (e.g.,
pe sonal skills, knowledge, echnologies, social connec ions)
o c ea e alue o hemsel es and o he s—jus like la ge
collec i es (Spoh e e al., 2008). Thus, e en an indi idual
can unc ion as a se ice sys em because hey engage in
mu ual se ice- o -se ice exchanges wi hin b oade ne -
wo ks (Va go & Lusch, 2011). A he same ime, supe o di-
na e se ice sys ems (e.g., i ms, neighbo hoods, o na ions)
a e hemsel es composed o in e dependen indi idual se -
ice sys ems in e ac ing and co-c ea ing alue.
The same a ionale applies o “se ice ecosys ems,”
de ined as nes ed cons ella ions o se ice sys ems (S o -
backa e al., 2016). Concep ually, “a se ice ecosys em may
be nes ed wi hin o be pa o a la ge sys em. Hence, se -
ice ecosys ems a e sys ems o sys ems in which he a ious
sys ems in e ac ” (S o backa e al. 2016, p. 3009). Schola s
equen ly compa e he e ms “se ice sys em” and “se ice
ecosys em,” no ing ha bo h e e o mul i-ac o a ange-
men s engaged in alue co-c ea ion (Wieland e al., 2012).
A key di e ence lies in he idea o analy ical zoom: se ice
ecosys ems emphasize b oade , eme gen ne wo ks o in e -
ac ing ac o s zooming ou om dyadic in e ac ions and dis-
c e e ansac ions o mo e complex ac o - o-ac o ne wo ks,
whe eas a se ice sys em can e e o any en i y—be i an
indi idual o collec i e—engaged in alue co-c ea ion ia
esou ce in eg a ion (Va go & Lusch, 2017; Poeppelbuss
e al., 2022).
By zooming ou , se ice esea ch has mo ed om s udy-
ing isola ed se ice sys ems (e.g., se ice i ms o se ice
deli e y sys ems) o analyzing he mo e dynamic, in e -
dependen con ex s o en i e se ice ecosys ems, in which
nume ous ac o s simul aneously co-c ea e alue (B ozo ic
& T egua, 2022). Because se ice ecosys ems “a e con-
s an ly adap ing o changing con ex ual equi emen s and
a e simul aneously c ea ing hese changing con ex s in he
p ocess” (Wieland e al., 2012, p. 15), small-scale and la ge-
scale shi s in ecosys em p ope ies can be unde s ood as
eme gen changes o phase ansi ions (Va go e al., 2023;
Polese e al., 2021).
When concep ualizing HISE, we hus d aw on ecen S-D
logic li e a u e, which p edominan ly employs se ice eco-
sys em e minology. Taking a sys ems pe spec i e on se ice
(Ba ile e al., 2016; Wieland e al., 2012) and unde s anding
se ice ecosys ems as sys ems o sys ems (S o backa e al.,
2016) has inspi ed a ious concep ualiza ions o agg ega ion
le els acco ding o S-D logic (see Table1).
Fo ins ance, Va go and Lusch (2017) dis inguish among
h ee le els o agg ega ion in se ice ecosys ems—mic o,
meso, and mac o—co esponding espec i ely o dyadic
exchanges (e.g., ansac ions, sha ing), an indus y o ma ke ,
and b oade socie al en i ies (e.g., local, na ional, o global
communi ies). Chandle and Va go (2011) likewise mo e
om dyads o iads and complex ne wo ks as uni s o analy-
sis, he eby cla i ying how se ice- o -se ice exchange can
be scaled up om a he di ec o mo e complex and indi ec
in e ac ions. They also in oduce he concep o a dynamic
me a laye (no le el) ha “ ep esen s [ he] e olu ion o hese
le els, which occu s simul aneously [o e ime]” (Chandle
& Va go, 2011, p. 41). In his sense, he me a laye “co -
e s all he le els o se ice- o -se ice exchanges such ha
hey oge he cons i u e se ice ecosys ems” (Chandle &
Va go, 2011, p. 44). S o backa e al. (2016) simila ly dis in-
guish h ee le els—mic o (ac o engagemen ), meso (se s o
ac o s, hei esou ces, engagemen pla o ms, and esou ce
in eg a ion pa e ns), and mac o ( he o e all ecosys em wi h
i s ins i u ional a angemen s). Ba ile e al. (2016) p opose a
“ i-le el app oach,” linking he se ice sys em concep om
se ice science (Maglio & Spoh e , 2008) wi h he se ice
ecosys em concep om S-D logic (Va go & Lusch, 2011).
Speci ically, hey illus a e how se ice sys ems, ne wo ked
se ice sys ems (also labeled as se ice ne wo ks), and se -
ice ecosys ems each ep esen dis inc ye in e wined ana-
ly ical le els in se ice esea ch. Vink e al. (2021) apply
he mic o-meso-mac o dis inc ion o a p ocess model o
se ice ecosys em design, designa ing he mic o le el as he
ocal ins ance o se ice ecosys em design, he meso le el
as encompassing bo h aligning and con lic ing design and
non-design p ocesses, and he mac o le el as he eme ging
ins i u ionalized pa e ns o alue co-c ea ion.
I is gene ally assumed ha ac o s can in e ac a all le -
els o agg ega ion and ha changes in one le el a ec he
o he s (Polese e al., 2021). Sa no e al. (2024), o exam-
ple, desc ibe a “domino e ec ” in which modi ica ions
wi hin one nes ed se ice ecosys em can p opaga e ou -
wa d o a b oade se ice ecosys em, o en beginning wi h
in e ac ions and adjus men s a he mic o le el and hen
“sp ead[ing] o he mac o le el.” Meynha d e al. (2016)
Elec onic Ma ke s (2025) 35:63 63 Page 6 o 27
Table 1 Agg ega ion le els acco ding o S-D logic
Se ice
ecosys em
le els
Va go and Lusch (2016), Va go
and Lusch (2017)
Chandle and Va go (2011) S o backa e al. (2016) Ba ile e al. (2016) Vink e al. (2021)
Mic o Indi idual and dyadic s uc u es
and ac i i ies in ol ed in dyadic
exchange (e.g., ansac ions,
sha ing)
Uni s o analysis a e dyads wi h
se ice- o -se ice exchange
among ac o s
Ac o engagemen (including
co-p oduc ion and alue-in-use
ac i i ies)
Se ice sys ems, which a e
dynamic con igu a ions o
esou ces (people, echnology,
in o ma ion, o ganiza ions) con-
nec ed in e nally and ex e nally
h ough alue p oposi ions
The ocal ins ance o se ice
ecosys em design as a eedback
loop o e lexi i y and e o ma-
ion (design) embedded wi hin
he p ocess o ep oduc ion
(non-design)
Meso Mid ange s uc u es and ac i i ies
(e.g., ma ke , indus y, b and
communi y, ca el)
Uni s o analysis a e iads wi h
se ice- o -se ice exchange
among dyads (e.g., wo dyads
o ac o s a and b and b and c,
wi h indi ec se ice- o -se ice
exchange be ween ac o a and c
h ough ac o b)
Se s o ac o s and hei esou ces
(including engagemen
pla o ms as in e media ies
be ween ac o s and e ol ing
esou ce in eg a ion pa e ns)
Ne wo ked se ice sys ems
(se ice ne wo ks) ha e ol e
h ough echnology, knowledge,
cul u e, business, and socie y
The in e play be ween bo h con-
lic ing and aligning design and
non-design p ocesses
Mac o B oade socie al s uc u es and
ac i i ies (e.g., socie y, na ional,
global, o local communi y)
Uni s o analysis a e complex ne -
wo ks wi h se ice- o -se ice
exchange among iads. When
complex ne wo ks success ully
ins i u ionalize esou ces, hey
become joined oge he as a
se ice ecosys em
Ecosys em and ins i u ional logic Se ice ecosys ems, which a e
ela i ely sel -con ained, sel -
adjus ing sys ems o esou ce-
in eg a ing ac o s, connec ed by
sha ed ins i u ional logics and
mu ual alue c ea ion h ough
hei se ice exchanges
The pa e ns o alue co-c ea ion
ha eme ge be ween design and
non-design p ocesses wi hin ac o
collec i es
Elec onic Ma ke s (2025) 35:63 Page 7 o 27 63
simila ly discuss bo om-up eme gence and op-down
ensla emen as dynamics in se ice ecosys ems: mac o-
le el p ope ies can eme ge om mic o-le el in e ac ions
in ways no ully de e mined by any one elemen , and hese
mac o-le el p ope ies can, in u n, eshape (“ensla e”)
indi idual (mic o-le el) elemen s.
Va go and Lusch (2016, p. 17) cau ion ha such dis inc-
ions o le els a e “ ela i e, a he han absolu e and hus
hese assignmen s a e somewha a bi a y” wi hin S-D logic.
Indeed, hey exis “as analy ical le els only and do no exis
independen ly o each o he . Ra he , hey ep esen pe -
spec i es ela ed o le els o agg ega ion” (Va go & Lusch,
2016, p. 18). I is also impo an o no con use le els o
agg ega ion (i.e., mic o-meso-mac o iews o mul i-ac o
ne wo ks) wi h le els o abs ac ion o building heo y
in esea ch (i.e., me a heo y, mid ange heo y, and mic o
heo y) (Va go & Lusch, 2017).
The key akeaway o ou concep ualiza ion o HISE is
ha “se ice ecosys ems” e e o se s o in e connec ed
ac o s and hei ela ionships, and analys s can zoom in
on dyadic ac o - o-ac o exchanges o zoom ou o exam-
ine en i e indus ies o socie ies. The ela ionships become
mo e complex a highe le els o abs ac ion, whe e no all
ac o s a e di ec ly connec ed anymo e (Chandle & Va go,
2011), and i is he esea che who “mus de e mine he el-
e an se ice ecosys em(s) and i s bounda ies o a pa icula
analysis” (Lusch e al., 2016, p. 2960).
I espec i e o how we de ine o cons ain a se ice eco-
sys em, all ac o s wi hin i engage in esou ce in eg a ion
(Va go & Lusch, 2008). This p ocess in ol es combining
bo h ope an esou ces (e.g., knowledge, skills, and abili ies)
and ope and esou ces (e.g., aw ma e ials o angible asse s)
o c ea e alue (Va go & Lusch, 2017). Resou ce in eg a ion
unde pins co e assump ions o S-D logic by emphasizing
ha e e y ac o in a se ice ecosys em is a esou ce in e-
g a o who ac i ely shapes alue co-c ea ion by combin-
ing and applying esou ces om a ious sou ces (Va go &
Lusch, 2016). A cen al dis inc ion he e is be ween ope and
esou ces, which a e gene ally passi e and equi e ex e nal
ac ion o become aluable (Cons an in & Lusch, 1994; Hun ,
2004), and ope an esou ces, which a e dynamic and knowl-
edge-based, o e ing he abili y o ac upon o he esou ces
o c ea e alue (Cons an in & Lusch, 1994; Hun , 2004). Fo
example, human skills and o ganiza ional capabili ies exem-
pli y ope an esou ces ha con e s a egic and compe i i e
ad an age (Va go & Lusch, 2016).
The se ice ecosys em pe spec i e u he highligh s
ha alue is co-c ea ed wi hin mul i-ac o sys ems ha
a e bound o sha ed ules, oles, no ms, and belie s, col-
lec i ely e med he ins i u ional a angemen s ha guide
esou ce in eg a ion and se ice exchange (Va go & Lusch,
2016; Vink e al., 2021). These ins i u ional no ms and
social ules p o ide a b oade con ex ha shapes how
alue co-c ea ion occu s by in luencing and s uc u ing he
in e ac ions be ween ac o s (O likowski, 1992; Walsham &
Han, 1991). Mo eo e , Lusch and Nambisan (2015) empha-
size ha all ac o s, including inanima e (i.e., ma e ial o
a i icial) agen s, ec ea e and al e na e hese ins i u ional
a angemen s h ough hei ac ions. Table2 summa izes
he undamen al concep s o S-D logic and p o ides de ini-
ions, which a e applied as a heo e ical lens o ou own
concep ualiza ion o HISE.
Recen esea ch has inc easingly examined how ech-
nology, pa icula ly AI, unc ions in se ice ecosys ems
(Kaa emo & Helkkula, 2024; Manse Payne e al., 2021;
Neuho e e al., 2021). AI-d i en echnologies a e ypi-
cally iewed as ope and esou ces ha can be in eg a ed
wi h human capabili ies o acili a e alue co-c ea ion on
highe le els (e.g., B eidbach & B odie, 2017). They also
enable complex in e ac ions among ne wo ks o human and
non-human ac o s, expanding oppo uni ies o alue c ea-
ion (S o backa e al., 2016). Fu he mo e, eme ging egu-
la ions such as he EU AI Ac (Eu opean Union, 2024) a e
implemen ed o o m new ins i u ional a angemen s ha
shape con empo a y se ice ecosys ems. While Va go and
Lusch (2017) an icipa ed he g owing impo ance o cogni-
i e compu ing and AI-powe ed sma se ice wi hin se ice
ecosys ems, hey la gely main ained a adi ional iew o
echnology as an ope and esou ce o e en black boxes any
echnological aspec s. In con as , wi hin he gi en con ex
o HISE, we conside echnology an ac i e pa icipan in
he se ice ecosys em—an app oach explo ed in mo e dep h
below.
Agency inin o ma ion sys ems esea ch
The IS discipline has adi ionally been subjec o he p i-
macy o human agency dominance (G asho & Recke ,
2023), whe e echnological a i ac s a e o en iewed solely
as passi ely used ools (Bai d & Ma uping, 2021) and hus
as ope and esou ces om an S-D logic pe spec i e. How-
e e , wi h he ad en o ad anced AI echnologies, he e
is a g owing ecogni ion o he agen ic capabili ies o AI
sys ems, leading o a eexamina ion o hei ole wi hin IS
esea ch and beyond.
Va ious li e a u e s eams, such as ac o -ne wo k heo y
(e.g., B aa & Sahay, 2004; Chiasson & Da idson, 2005;
Hanse h e al., 2006; Lamb & Kling, 2003; Sco & Wag-
ne , 2003), socioma e iali y (e.g., Cecez-Kecmano ic
e al., 2014; Leona di, 2011), and agen -based compu ing
(e.g., B enne e al., 1998; Mille & Pa asu aman, 2007;
Russell, 2019), posi ha human agen s a e in mu ual
ela ionships wi h agen ic in o ma ion echnology (IT)
a i ac s. These ela ionships o m cons an ly ( e-)eme g-
ing s uc u es wi h hemsel es and hei con ex in p ac-
ice (O likowski, 2000). In his con ex , AI sys ems a e
Elec onic Ma ke s (2025) 35:63 63 Page 8 o 27
inc easingly iewed as possessing agency— he capaci y
o ac au onomously and in luence ou comes— he eby
becoming ac i e pa icipan s in o ganiza ional p ocesses.
Bai d and Ma uping (2021) call o a heo e ical delega-
ion amewo k ha s esses he impo ance o how igh s,
esponsibili ies, and coo dina ion occu be ween human
and machine agen s and how his ela ionship e ol es (e.g.,
Akinola e al., 2018; Klein e al., 2006; Leana, 1986; Ribes
e al., 2013). This shi owa ds acknowledging ha IT a i-
ac s, pa icula ly AI sys ems, ha e agency leads o a b oad-
ened iew on heo izing he ela ionship be ween echnol-
ogy and humans. By ecognizing he oscilla ing na u e o
in e ac ions be ween agen ic IT a i ac s and humans, new
s uc u es a ise, each depending on cons i u ing ac o s o
ei he human o IT a i ac a ibu es.
Bai d and Ma uping (2021) concep ualize h ee main
a ibu es o bo h human agen s and agen ic IT a i ac s:
endowmen s, p e e ences, and oles, which p o ide a concep-
ual basis o unde s anding he ecen shi om adi ional
agency concep ualiza ion owa ds a dynamic delega ion-
o ien ed pe spec i e. In e es ingly, all h ee agency- ele an
a ibu es a e highly cohesi e wi h S-D logic.
F om an S-D logic pe spec i e, endowmen s a e he
ac o s’ ope and (e.g., da a) and ope an esou ces (e.g.,
knowledge and hinking capabili ies). The skills and knowl-
edge a ibu ed o human-a i ac dyads equi e a ce ain
Table 2 Fundamen al concep s o S-D logic
S-D logic concep De ini ions and ele an axioms o S-D logic Ci a ion
Se ice Se ice is he “applica ion o ope an esou ces (skills and knowledge)” o ano he
pa y (i.e., ac o ). “Se ice is exchanged o se ice.”
“Se ice p o ision implies he ongoing combina ion o esou ces, h ough in eg a-
ion, and hei applica ion, d i en by ope an esou ces — he ac i i ies o ac o s.”
“Se ice is he undamen al basis o exchange.” (Axiom 1)
Va go and Lusch (2008), p. 6
Va go and Lusch (2011), p. 184
Va go and Lusch (2016), p. 18
Ac o “All social and economic ac o s a e esou ce in eg a o s.”
“Ac o s need o be iewed no only as humans, bu also as machines/ echnologies,
o collec ions o humans and machines/
echnologies, including o ganiza ions.”
“[…] s uc u es connec [mul iple] ac o s and p o ide hei con ex and become
ac o s hemsel es. […] In he se ice-science amewo k, hey a e se ice sys-
ems.”
Va go and Lusch (2008), p. 6
S o backa e al. (2016), p. 3010
Va go and Lusch (2011, p. 186)
Se ice ecosys em “A se ice ecosys em is a spon aneously sensing and esponding spa ial and em-
po al s uc u e o la gely loosely coupled, alue-p oposing social and economic
ac o s in e ac ing h ough ins i u ions, echnology, and language o (1) co-p o-
duce se ice o e ings, (2) engage in mu ual se ice p o ision, and (3) co-c ea e
alue.”
“[A] ela i ely sel -con ained, sel -adjus ing sys em o esou ce-in eg a ing ac o s
connec ed by sha ed ins i u ional a angemen s and mu ual alue c ea ion h ough
se ice exchange”
Va go and Lusch (2011), p. 185
Va go and Lusch (2016), pp. 10
Ope and esou ces Resou ces “ ha equi e some ac ion o be pe o med on hem o ha e alue (e.g.
na u al esou ces).”
Va go and Lusch (2011), p. 184
Ope an esou ces Resou ces “ ha can be used o ac (e.g. human skills and knowledge).” Va go and Lusch (2011), p. 184
Value co-c ea ion “The p ocesses and ac i i ies ha unde lie esou ce in eg a ion and inco po a e
di e en ac o oles in he se ice ecosys em.”
“Value is coc ea ed by mul iple ac o s, always including he bene icia y.” (Axiom
2)
Lusch and Nambisan (2015), p. 162
Va go and Lusch (2016), p. 18
Value Value is an “eme gen , posi i ely o nega i ely alenced change in he well-being
o iabili y o a pa icula sys em/ac o .”
Value “is iewed as an imp o emen in a sys em as de e mined by he sys em o
by he sys em’s abili y o adap o an en i onmen . In o he wo ds, alue can be
concep ualized as imp o ed sys em iabili y.”
“[…] alue occu s when he o e ing is use ul o he cus ome o bene icia y
( alue-in-use), and his is always in a pa icula con ex .”
“Value is always uniquely and phenomenologically de e mined by he bene icia y.”
(Axiom 4)
Va go and Lusch, (2018), p. 740
Wieland e al. (2012), p. 17
Lusch and Nambisan (2015), p. 159
Va go and Lusch (2016), p. 18
Ins i u ions and ins i-
u ional a ange-
men s
Ins i u ions a e “humanly de ised ules, no ms, and belie s ha enable and con-
s ain ac ion and make social li e p edic able and meaning ul.”
Ins i u ional a angemen s a e “se s o in e ela ed ins i u ions (some imes e e ed
o as ‘ins i u ional logics’).”
“Value co-c ea ion is coo dina ed h ough ac o -gene a ed ins i u ions and ins i u-
ional a angemen s.” (Axiom 5)
Va go and Lusch (2016), p. 6
Va go and Lusch (2016), p. 11
Va go and Lusch (2016), p. 18
Elec onic Ma ke s (2025) 35:63 Page 15 o 27 63
Semi‑au onomous d i ing
Gi en he apid ad ancemen o AI echnologies in he au o-
mo i e indus y, semi-au onomous d i ing se es as a com-
pelling applica ion o he HISE amewo k. The in eg a ion
o AI in o he mobili y landscape demons a es he dynamic
na u e o esou ces, as da a sou ces, senso echnologies,
and AI algo i hms con inuously e ol e and adap o new
condi ions. This in eg a ion ans o ms alue co-c ea ion by
enhancing sa e y, inc easing d i ing e iciency, and en ich-
ing he o e all d i ing expe ience. Mul iple ac o s wi hin
he ecosys em o semi-au onomous d i ing collabo a e o
co-c ea e, implemen , and e ine hyb id in elligen se ices
ha enable sa e and mo e e icien anspo a ion.
In his scena io, a d i e ope a es a semi-au onomous
ehicle equipped wi h AI-powe ed ad anced d i e assis-
ance sys ems (ADAS), including ea u es such as adap i e
c uise con ol, lane-keeping assis ance, and collision a oid-
ance. The d i e , as a human-AI hyb id, con ibu es human
expe ise insi ua ional awa eness, decision-making, and
o e all ehicle con ol, pa icula ly in a e e en s—so-called
“edge cases.” Simul aneously, he AI sys em p o ides eal-
ime da a analysis and decision suppo by p ocessing in o -
ma ion om senso s, came as, and ex e nal da a sou ces
such as mapping, a ic, and wea he upda es.
The dynamic in e play be ween human and a i icial
agency is e iden : o example, while he AI sys em manages
ou ine asks such as lane-keeping and adap i e c uise con-
ol, he d i e e ains con ol du ing c i ical o unexpec ed
a ic condi ions. The d i e elies on expe ise o na iga e
complex en i onmen s and makes high-le el decisions,
such as in e p e ing ambiguous oad signs o esponding
o un o eseen si ua ions. Meanwhile, he AI sys em assis s
wi h eal- ime adjus men s, like main aining a sa e dis ance
om o he ehicles, adhe ing o lanes, o applying au oma ic
b aking o a oid collisions. In ex eme si ua ions, he AI
sys em migh ale he human d i e o ake o e con ol, o
con e sely, i may execu e eme gency maneu e s au ono-
mously. This con inuous nego ia ion and econ igu a ion
o agency be ween he human d i e and he AI-powe ed
ADAS exempli ies he concep o con igu ed agency in
human-AI hyb ids, while i can be assumed ha he con-
inuum on which humans alloca e asks o he ADAS a ies
acco ding o indi idual p e e ences and he speci ic con ex .
An in e es ing aspec o his scena io is ha ac ual d i ing
can be unde s ood as a “sel -se ice,” whe e a human ac o
bo h p o ides and bene i s om he se ice o con olling
he ehicle, while passi e ac o s (passenge s) a e bene icia -
ies o he mobili y se ice. This highligh s he dual ole o
he d i e as bo h a se ice p o ide (ope a ing he ehicle)
and a bene icia y (gaining mobili y), unde sco ing he com-
plexi y o se ice in e ac ions wi hin he HISE, whe e oles
can o e lap and shi depending on he con ex .
Mo eo e , he in e ac ions be ween ac o s wi hin he
HISE ex end beyond he d i e and he ehicle’s AI sys ems.
The d i e and ehicle in e ac wi h o he human-AI hyb ids
(d i e s using ADAS), human ac o s (passenge s, pedes i-
ans, d i e s wi hou AI assis ance), and po en ially AI ac o s
( ully au onomous ehicles). Fo example, he ehicle may
communica e wi h nea by au onomous ehicles o coo di-
na e mo emen s and op imize a ic low. This necessi a es
in eg a ing esou ces such as sha ed oad in as uc u e,
communica ion p o ocols, and da a exchange mechanisms.
Resou ces in his scena io a e dynamic and modula ,
including physical in as uc u e ( oads, a ic signs, gas
s a ions), senso da a, mapping in o ma ion, and AI algo-
i hms. The con inuous upda ing o mapping da a, a ic
in o ma ion, and so wa e upda es o AI sys ems exempli-
ies he dynamic na u e o esou ces, which ypically adhe e
o de ined s uc u es (e.g., speci ied da a o ma s, APIs,
access igh s o in- ehicle da a, e c.) ha acili a e seamless
in e ope abili y. Ac o s wi hin he ecosys em can selec i ely
deploy di e en combina ions o hese esou ces depend-
ing on si ua ional needs, op imizing join alue c ea ion.
Fo ins ance, he ehicle’s ope a ing sys em consolida es
da a om senso s and ex e nal sou ces o assess he d i -
ing en i onmen , an icipa e po en ial haza ds, and op imize
d i ing pe o mance. Addi ionally, da a ma ke places may
enable hi d pa ies, such as insu ance companies o main-
enance p o ide s, o con ibu e addi ional se ices, enhanc-
ing he ecosys em’s o e all alue. O e all, he p e ailing
esou ce s uc u e empowe s ac o s o u ilize esou ces o
make well-in o med decisions, adap o e ol ing condi ions,
and achie e supe io ou comes in e ms o sa e y, e iciency,
and well-being. The dynamic na u e o esou ces is u he
ecognized h ough he con inuous ad ancemen s in AI
echnologies, senso capabili ies, and communica ion in a-
s uc u es. Fo example, new algo i hms o be e objec
ecogni ion o imp o ed decision-making in complex a ic
scena ios may be de eloped and in eg a ed in o he ADAS.
Simila ly, upda es o mapping in o ma ion o a ic da a
sys ems e lec changes in oad condi ions, cons uc ion,
o a ic pa e ns, equi ing he human-AI hyb id o adap
acco dingly.
Ins i u ional a angemen s play a c ucial ole in shaping
he beha io s and in e ac ions o ac o s wi hin his HISE
scena io. Go e nmen egula ions, indus y s anda ds,
and sa e y guidelines—such as he SAE le els o d i ing
au oma ion (Socie y o Au omo i e Enginee s, 2021) and
associa ed sa e y s anda ds—go e n he de elopmen and
deploymen o AI-d i en capabili ies, ensu ing sa e y while
p omo ing inno a ion. Fo example, egula ions may speci y
he equi ed sa e y ea u es o ADAS o se s anda ds con-
ce ning da a p i acy and cybe secu i y in connec ed ehi-
cles. Da a-sha ing ag eemen s and p o ocols enable smoo h
da a in eg a ion om di e se sou ces, os e ing e ec i e
Elec onic Ma ke s (2025) 35:63 63 Page 16 o 27
collabo a ion among ac o s while e lec ing da a so e eign y
p inciples. In o mal ne wo ks and collabo a ions be ween
manu ac u e s, echnology p o ide s, and se ice p o ide s
con ibu e o knowledge sha ing and he e olu ion o bes
p ac ices wi hin he ecosys em.
The in e play be ween indi idual beha io s, o ganiza-
ional esponses, and egula o y ac ions exempli ies how
di e en le els wi hin he HISE in luence each o he . Fo
ins ance, inciden s, whe e d i e s o semi-au onomous ehi-
cles misused he echnology by engaging in ac i i ies like
sleeping while d i ing, ha e p omp ed egula o y bodies o
sc u inize and demand changes om manu ac u e s. This has
led o adjus men s in he design and unc ionali y o ADAS
o limi inapp op ia e use and en o ce sa e agency con igu-
a ions. Tesla, o example, has been equi ed o implemen
d i e moni o ing sys ems ha de ec ina en i eness and
issue ale s o e en disable au onomous ea u es i neces-
sa y. Such dynamics illus a e how ac o s ac oss di e en
le els o he HISE in e ac o en o ce ins i u ional a ange-
men s, ensu ing ha human esponsibili y is app op ia ely
main ained wi hin agency con igu a ions. These egula o y
in e en ions ep esen a sel - egula ing e ec o he ecosys-
em, whe e mac o-le el au ho i ies en o ce ins i u ional ules
ha necessi a e ac o s o adjus hei echnologies, ul ima ely
in luencing mic o-le el beha io s.
Elde ly ca e
Elde ly ca e ep esen s ano he illus a i e case o apply-
ing he HISE amewo k, as in eg a ing AI echnolo-
gies in o he human ca e se ice sec o can signi ican ly
enhance he quali y o li e o olde adul s. By blending
human expe ise wi h AI capabili ies, AI-augmen ed
elde ly ca e se ices imp o e alue-in-use h ough pe son-
alized ca e plans, e icien moni o ing, and, ul ima ely, be -
e well-being ou comes. Mul iple ac o s wi hin he elde ly
ca e se ice ecosys em collabo a e o design, implemen ,
and e ine hyb id in elligen se ice o e ings ha p omo e
op imal ca e and suppo .
In his scena io, a ca egi e and a ca e coo dina o col-
labo a e wi h an AI-based heal h and acili y managemen
sys em o p o ide comp ehensi e ca e o an olde adul , he
p ima y bene icia y. The ca egi e and ca e coo dina o a e
conside ed human-AI hyb ids because hey ac i ely con ig-
u e and in eg a e bo h human and a i icial agencies in hei
oles. The ca egi e con ibu es human empa hy, compas-
sion, and ailo ed ca e, while he ca e coo dina o o e s
expe ise in ca e managemen and coo dina ion o suppo
se ices. The AI sys em assis s by ga he ing and analyzing
da a om a ious sou ces, including wea ables, sma home
de ices, and heal h eco ds, gene a ing pe sonalized ec-
ommenda ions o daily ou ines, medica ion managemen ,
ea ly de ec ion o po en ial heal h issues, and imp o ing
he e iciency o ca e managemen p ocesses in he acil-
i y. In domains such as elde ly ca e, which ace sho ages
o highly skilled wo ke s, his ees up ime o he s a o
ocus on hei co e compe encies in human in e ac ion wi h
he bene icia ies ins ead o edious asks such as epo ing
medicine in ake.
The dynamic in e play be ween human and a i icial
agencies is e iden as he ca egi e and ca e coo dina o
nego ia e he con igu a ion o agencies o op imized ask
pe o mance. Fo example, he ca egi e may ely on hei
human expe ise o p o ide emo ional suppo and ecognize
sub le changes in he pa ien ’s beha io ha equi e a en-
ion. Simul aneously, he AI sys em moni o s heal h indica-
o s in eal ime, analyzes pa e ns in he da a, and p o ides
ale s o sugges ions, enabling p oac i e in e en ions. This
collabo a ion enhances he quali y and esponsi eness o
ca e, illus a ing he con igu ed agency wi hin he human-
AI hyb id.
Pa ien s wi hin he elde ly ca e se ice ecosys em can be
desc ibed as human ac o s who do no necessa ily become
hyb ids. While hey in e ac wi h AI sys ems indi ec ly
h ough he se ices p o ided, hey ypically do no ac i ely
con igu e agency be ween hemsel es and AI sys ems. How-
e e , hey do in eg a e esou ces by using wea able de ices
and heal h moni o ing senso s, which collec da a essen-
ial o he AI sys em’s unc ioning. This pa icipa ion is
c ucial o esou ce in eg a ion wi hin he HISE, al hough
he pa ien s hemsel es do no con igu e agency wi h AI
sys ems. This enables he o he ac o s o u ilize esou ces
o make in o med decisions e ec i ely, adap o changing
needs, and achie e be e ou comes in elde ly ca e and sup-
po han could be achie ed wi hou he in ol emen o AI.
The se ice ecosys em could also include esou ces ha
allow o in eg a ion wi h adjacen se ice ecosys ems, such
as an elec onic heal h eco d ha enables sha ing o heal h-
ela ed da a wi h hi d-pa y heal h p o essionals.
Resou ces in his scena io include ope an esou ces like
he ca egi e ’s skills and knowledge, and ope and esou ces
like wea able de ices, heal h moni o ing senso s, AI algo-
i hms, and communica ion pla o ms. These esou ces
a e dynamic, con inuously adap ing o new echnologies,
pa ien needs, and medical ad ancemen s. Fo example,
new wea able de ices may o e addi ional heal h me ics,
o AI algo i hms may be upda ed based on he la es medical
esea ch, equi ing ac o s o adap and in eg a e hese e ol -
ing esou ces in o he ca e p ocess.
In his scena io in ol ing ulne able ac o s, he mac o-
le el ins i u ional a angemen s play a signi ican ole in
shaping he beha io and in e ac ions o ac o s wi hin he
HISE. Regula ions and s anda ds o p i acy and secu i y
go e n he handling o sensi i e pe sonal heal h in o ma ion,
ensu ing compliance wi h laws such as HIPAA in he USA
o GDPR in Eu ope. In addi ion, p o essional guidelines,
Elec onic Ma ke s (2025) 35:63 Page 17 o 27 63
quali y managemen p ocesses, and ce i ica ion equi e-
men s o ca egi e s and ca e coo dina o s main ain high
s anda ds o ca e and se ice quali y. Fo ins ance, he e
may be p o ocols o esponding o speci ic heal h ale s
gene a ed by he AI sys em, ensu ing e hical and e ec i e
in e en ions. In o mal knowledge-sha ing ne wo ks, such
as online o ums, communi y g oups, and wo kshops among
ca egi e s, heal hca e p o essionals, and se ice p o ide s,
can also shape he ecosys em a ound elde ly ca e se ices.
These ne wo ks os e he sha ing o bes p ac ices, inno a-
ions, and expe iences, which d i e u he ad ances in he
ield and imp o e he quali y o ca e o olde adul s ac oss
human agency, a i icial agency, and agency con igu a ion.
Sus ainable coding
The so wa e indus y is esponsible o app oxima ely 1.8%
o 2.8% o global g eenhouse gas emissions (F ei ag e al.,
2021). The e o e, add essing sus ainabili y in so wa e
enginee ing has become c i ical, wi h a ious ini ia i es
aiming o educe emissions and p omo e en i onmen ally
iendly p ac ices (Ma e al., 2022; Naumann e al., 2011).
Fo example, minimizing he ca bon oo p in o so wa e
can be achie ed by educing da a s o age, da abase eques s,
and API calls when de eloping so wa e. Ca bon- iendly
p og amming ep esen s a scena io o hyb id in elligen se -
ice whe e so wa e enginee s collabo a e wi h AI sys ems
o de elop sus ainable so wa e, he eby educing he ca bon
oo p in associa ed wi h so wa e applica ions. This sce-
na io illus a es how hyb id in elligence can ans o m and
enhance alue co-c ea ion p ocesses by enabling he de el-
opmen o en i onmen ally sus ainable so wa e solu ions.
In his con ex , a human-AI hyb id is o med by a so -
wa e enginee wo king alongside an AI-based coding assis-
an , such as an ins ance o Cha GPT o Gi Hub Copilo
speci ically ained o de eloping ca bon- iendly code
o clien applica ions. The so wa e enginee con ibu es
human agency h ough c ea i i y, equi emen s enginee ing,
and p og amming expe ise. They unde s and he so wa e’s
concep , use equi emen s, and design speci ica ions, b ing-
ing c i ical hinking and p oblem-sol ing skills o he de el-
opmen p ocess. Meanwhile, he AI sys em p o ides a i i-
cial agency by gene a ing ca bon- iendly code, op imizing
algo i hms o ene gy e iciency, and iden i ying pa e ns
ha con ibu e o unnecessa y emissions.
The con igu a ion o human and a i icial agency is e i-
den as asks a e alloca ed based on si ua ional needs and
expe ise. Fo ins ance, he so wa e enginee may w i e ini-
ial code s uc u es and algo i hms, le e aging hei c ea i -
i y and unde s anding o he so wa e’s in ended unc ion-
ali y. Simul aneously, he AI sys em analyzes he code o
po en ial ine iciencies, sugges s al e na i e coding p ac ices
ha educe compu a ional esou ce usage, and op imizes he
code o sus ainabili y wi hou comp omising pe o mance
o secu i y. Whe he he AI sys em con ibu es in a sequen-
ial, pa allelized, o i e a i e p ocess can be con igu ed lex-
ibly, accoun ing o p ojec -speci ic goals, p e e ences, and
esou ce a ailabili ies.
Fo example, du ing he de elopmen p ocess, he so -
wa e enginee migh design a da a p ocessing module ha
handles la ge da ase s. The AI assis an could p oac i ely
analyze he expec ed load and da a access pa e ns and,
he e o e, ecommend mo e e icien da a s uc u es o algo-
i hms ha equi e ewe compu a ional esou ces, he eby
educing ene gy consump ion. The AI sys em migh also
au onomously swi ch he applica ion om a adi ional
ela ional da abase o a mo e e icien NoSQL da abase like
MongoDB ha o e s be e pe o mance and lowe ene gy
consump ion o he gi en applica ion’s uns uc u ed da a
needs— hus aligning he so wa e in as uc u e wi h sus-
ainabili y goals. In a hyb id in elligen se ice in so wa e
de elopmen , he human-AI hyb id will in e ac wi h o he
hyb id o human ac o s in he so wa e de elopmen p ocess,
and he bene icia y (clien ) migh also use human, hyb id, o
e en AI ac o s, o example, ope o m a quali y e alua ion
on he de eloped code.
Resou ces in his scena io include ope an esou ces such
as he so wa e enginee ’s skills, knowledge, and compe en-
cies, and ope and esou ces like p og amming languages,
de elopmen en i onmen s, sou ce code edi o s, and access
igh s. Mo eo e , he AI sys em u ilizes ex ensi e, g owing
da ase s o coding p ac ices and sus ainabili y me ics o p o-
ide eal- ime eedback and op imiza ion sugges ions. These
esou ces a e dynamic, cons an ly e ol ing wi h echnologi-
cal ad ances and changing indus y s anda ds, as well as
wi h expe ience in p o iding his hyb id in elligen se ice.
Fo example, by ecei ing eedback om he AI sys em on
sus ainable so wa e de elopmen p ac ices, he so wa e
enginee may gain knowledge o u u e p ojec s ha allow
hem o adop hese p ac ices by hemsel es o imp o e hei
wo k low in collabo a ion wi h he AI sys em. Addi ional
esou ce s uc u es suppo his co-c ea ion p ocess by ena-
bling a luid exchange and in eg a ion o esou ces. S and-
a dized p og amming languages, code eposi o ies, and
de elopmen ools acili a e collabo a ion be ween human
enginee s and AI sys ems. Fo ins ance, Gi Hub Copilo
in eg a es seamlessly in o Visual S udio Code, c ea ing a
symbio ic de elopmen en i onmen whe e bo h human
enginee s and AI sys ems can access and collabo a e on so -
wa e sou ce code in eal ime, mo ing beyond he adi ional
ole o AI sys ems as a ool by enac ing hei own agency.
Ins i u ional a angemen s guide beha io s and in e -
ac ions wi hin he sus ainable coding HISE. Mac o-le el
so wa e indus y s anda ds, cybe secu i y, and da a secu-
i y amewo ks ensu e ha code mee s quali y and secu i y
equi emen s. En i onmen al egula ions and o ganiza ional
Elec onic Ma ke s (2025) 35:63 63 Page 18 o 27
policies may manda e sus ainabili y p ac ices, in luencing
so wa e de elopmen app oaches. In ellec ual p ope y
laws and licensing ag eemen s a ec he use o AI-gene a ed
code, ensu ing compliance and e hical conside a ions in he
collabo a ion o human and AI ac o s. Da a secu i y is also
pa amoun ; sensi i e clien da a appea ing in speci ica ions
mus be p o ec ed and explici ly excluded om u he ain-
ing o AI models o comply wi h p i acy egula ions.
P ecision ag icul u e
P ecision ag icul u e se es as an illus a i e case o HISE,
demons a ing he alue o hyb id in elligence in ans o m-
ing adi ional a ming p ac ices h ough he in eg a ion o
AI echnologies. The ag icul u al sec o has been a ela i ely
ea ly adop e o AI echnologies compa ed o o he domains.
By u ilizing da a-d i en insigh s and ad anced analy ics,
p ecision ag icul u e enhances alue co-c ea ion, e.g., by
op imizing esou ce u iliza ion, imp o ing c op yields, and
p omo ing sus ainable a ming p ac ices. The a ious ac o s
in he p ecision ag icul u e HISE wo k oge he o deli e
high-quali y ag icul u al p oduc s, le e aging hyb id in el-
ligen se ice o e ings ha acili a e e ec i e sha ed deci-
sion-making and esou ce alloca ion be ween hese ac o s.
In his scena io, human-AI hyb ids a e o med by a a me
and an ag onomis , espec i ely, collabo a ing wi h AI-based
decision suppo sys ems o manage a ield o c ops. The
a me p o ides human agency h ough local knowledge o
he land, unde s anding o c op his o y, and p ac ical a m-
ing expe ience. The ag onomis con ibu es expe ise in c op
managemen , soil science, and pes con ol s a egies. The
AI sys ems o e a i icial agency by in eg a ing da a om
mul iple sou ces—such as soil senso s, wea he his o y and
o ecas s, d one image y, and his o ical c op da a— o gene -
a e eal- ime ecommenda ions o op imal i iga ion sched-
ules, e ilize applica ion a es, o pes managemen in e -
en ions. Bo h o hese human-AI hyb ids no only in e ac
wi h each o he bu also wi h o he human (e.g., consume s,
ield wo ke s), hyb id (e.g., semi-au onomously ope a ed
ield machine y), o AI ac o s (e.g., au oma ed in en o y
managemen and p ocu emen sys ems) wi hin he HISE.
Fo ins ance, a dis ibu ion cen e o ag icul u al p oduc s
wi h an AI-based in en o y managemen and demand o e-
cas ing sys em could connec he hyb id a me wi h he
shops o consume s. This AI sys em au onomously o de s
ag icul u al p oduc s based on eal- ime da a, ensu ing a
seamless supply chain. Thus, i assumes he ole o an AI
ac o minimizing was e and ensu ing consume s’ access o
esh p oduc s. Human consume s, he end-use s o ag icul-
u al p oduc s, bene i om he high-quali y, sus ainably p o-
duced goods esul ing om his hyb id in elligen se ice.
The con igu a ion o human and a i icial agencies is e i-
den as asks a e dynamically alloca ed based on expe ise,
si ua ional demands, and le els o g anula i y. Fo exam-
ple, he a me o ag onomis may use hei judgmen based
on hei human expe ise o in e p e AI-gene a ed insigh s
o de e mine he app op ia e c op o a ion plan and soil
ea men wi hin he con ex o hei speci ic en i onmen ,
making s a egic decisions ha conside ac o s beyond he
analysis o he da a a ailable, such as ma ke condi ions o
communi y p ac ices. Simila ly, he AI sys em migh sug-
ges an op imal i iga ion schedule as a eal- ime ecommen-
da ion based on soil mois u e da a and wea he o ecas s,
bu he a me may adjus his ecommenda ion based on
knowledge o local wa e a ailabili y o i iga ion equip-
men cons ain s.
Resou ces in eg a ed by he ac o s in he p ecision ag i-
cul u e HISE include ope an esou ces like he knowledge
and expe ise o he a me and ag onomis , and ope and
esou ces such as ag icul u al machine y, IoT de ices,
emo e sensing da a, and AI algo i hms. The modula s uc-
u e o hese esou ces acili a es e ec i e esou ce in eg a-
ion by s anda dizing, e.g., da a o ma s, communica ion
p o ocols, and au ho iza ion sys ems. Fo example, an in e-
g a ed pla o m migh combine soil mois u e senso s, sa el-
li e image y, and AI-based models o moni o c op heal h,
p edic yield, and op imize esou ce use. This s anda d-
ized s uc u e allows u ilizing esou ces e ec i ely, mak-
ing in o med decisions, adap ing o changing condi ions,
and achie ing be e ou comes in e ms o p oduc i i y and
sus ainabili y. Fu he mo e, hese esou ces a e dynamic,
adap ing o con inuous ad ancemen s in senso echnolo-
gies, AI algo i hms, and da a analy ics. Fo example, new
senso echnologies may p o ide mo e accu a e soil mois-
u e eadings, o AI algo i hms may imp o e in p edic ing
pes in es a ions. As new da a sou ces become a ailable and
algo i hms imp o e, human-AI hyb ids mus adap hei
p ac ices o inco po a e hese de elopmen s, ensu ing ha
a ming emains esponsi e o en i onmen al condi ions and
echnological inno a ions.
Ins i u ional a angemen s play a c i ical ole in guiding
he beha io s and in e ac ions o ac o s wi hin he p ecision
ag icul u e HISE. Fo example, go e nmen subsidies and
ag icul u al policies (on he mac o le el) may encou age he
adop ion o sus ainable a ming p ac ices and AI echnolo-
gies such as a iable a e e iliza ion, p omo ing en i on-
men al s ewa dship and esou ce conse a ion by indi idual
ac o s (on he mic o le el). Indus y s anda ds o da a sha -
ing and in e ope abili y os e collabo a ion among di e se
ac o s in he se ice ecosys em, ensu ing ha da a om
di e en de ices and IT sys ems can be in eg a ed e ec-
i ely. En i onmen al egula ions in luence esou ce u iliza-
ion, pes managemen p ac ices, and he use o e ilize s
and pes icides, shaping he decisions made by human-AI
hyb ids. Fo example, compliance wi h en i onmen al egu-
la ions may equi e he a me and ag onomis o limi he
Elec onic Ma ke s (2025) 35:63 Page 19 o 27 63
use o ce ain chemicals. The AI sys em can assis by sug-
ges ing al e na i e pes managemen s a egies o op imizing
he applica ion o e ilize o mee egula o y equi emen s
while main aining c op heal h.
IT cus ome suppo
As IT se ice managemen (ITSM) g ows in complexi y,
o ganiza ions inc easingly s uggle wi h handling suppo
eques s e icien ly. The ising olume o se ice eques s,
he agmen a ion o knowledge ac oss dis ibu ed eams,
and wo k o ce u no e make i di icul o main ain high
se ice quali y. AI-powe ed au oma ion o e s po en ial solu-
ions by enhancing decision-making and op imizing icke
esolu ion, ye i s isola ed deploymen o en lacks adap -
abili y, explainabili y, and us among IT se ice employ-
ees. Resea ch analyzing o e 17,000 IT suppo icke s has
shown ha AI o en misclassi ies cases in edge scena ios,
equi ing human in e en ion o ensu e accu acy and con-
inuous lea ning (Li e al., 2024). This unde sco es he need
o a s uc u ed amewo k ha concep ualizes how hyb id
in elligence should be designed and managed in se ice
ecosys ems. The HISE amewo k p o ides his ounda ion
by de ining how human ac o s, AI ac o s, and human-AI
hyb ids in e ac in dynamic, co-e ol ing se ice en i on-
men s, enabling seamless alue co-c ea ion h ough hyb id
in elligen se ice. The HISE amewo k concep ualizes
se ice ecosys ems as dynamic ne wo ks whe e human, AI,
and human-AI hyb id ac o s collabo a e o op imize ask
pe o mance. The hyb id in elligen se ice suppo (HISS)
model (Reinha d e al., 2023) ins an ia es he HISE ame-
wo k by demons a ing how hyb id in elligence enhances
ope a ional scalabili y, adap abili y, and knowledge e en-
ion in IT cus ome suppo . Unlike adi ional au oma ion
models, HISE’s si ua ionally con igu ed agency enables
IT se ice agen s o con inuously nego ia e ask alloca ion
wi h AI-d i en decision suppo sys ems. Depending on he
complexi y and u gency o a icke , AI may ake he lead in
au oma ed issue classi ica ion, while human agen s in e ene
o high-s akes decision-making. Fo example, human-AI
hyb ids in ITSM nego ia e agency con igu a ions con inu-
ously, de e mining he op imal mix o AI au oma ion and
human expe ise o each se ice ins ance.
A key con ibu ion o HISE o IT se ice managemen is
i s s uc u ed iew o agency con igu a ions. ITSM in ol es
a ious ypes o se ice in e ac ions, anging om simple,
epea able asks ha can be ully au oma ed o complex,
high-s akes decisions ha equi e human expe ise. The
HISE amewo k p o ides a concep ual lens o unde s and-
ing how agency is dynamically dis ibu ed ac oss human and
AI ac o s. In he HISS scena io, ou ine se ice eques s—
such as passwo d ese s o basic oubleshoo ing—can be
handled by AI ac o s h ough cha bo s and au oma ed icke
classi ica ion. Mo e complex issues—such as diagnosing a
p e iously unseen ne wo k ou age o esol ing mul i-sys em
in eg a ion ailu es— equi e human expe ise, whe e human
agen s ake he lead. Howe e , HISE emphasizes ha many
se ice in e ac ions all be ween hese ex emes, equi ing
human-AI hyb id agency. In HISS, human agen s collabo a e
wi h AI-powe ed decision-suppo sys ems ha ecommend
solu ions, e ie e knowledge om pas inciden s, and high-
ligh po en ially ele an oubleshoo ing s eps. By le e ag-
ing he HISE amewo k’s iadic in e ac ion model, HISS
ac o s dynamically shi be ween human, AI, and hyb id
oles o op imize se ice e iciency.
Beyond s uc u ing agency con igu a ions, HISE cla i-
ies how esou ces a e in eg a ed and mobilized in hyb id
in elligence en i onmen s. IT se ice managemen elies on
a ious o ms o da a esou ces, including his o ical icke
logs, se ice knowledge bases, sys em moni o ing da a, and
eal- ime use eedback. AI models in HISS p ocess his da a
o classi y issues, p edic esolu ion pa hways, and sugges
solu ions based on p io cases. Howe e , HISE highligh s
ha da a alone is no enough— esou ce in eg a ion in hyb id
in elligence en i onmen s equi es con inuous knowledge
alida ion, e inemen , and adap a ion by human ac o s. IT
cus ome suppo applies his p inciple by inco po a ing
s uc u ed eedback loops, whe e se ice agen s e iew AI-
gene a ed ecommenda ions, co ec e o s, and lag knowl-
edge gaps ha equi e human-d i en upda es. O e ime,
his co-c ea ion p ocess shapes a sha ed knowledge base,
imp o ing AI p edic ions while ein o cing human expe ise.
This ensu es ha AI models e ol e in alignmen wi h o gan-
iza ional bes p ac ices and se ice-speci ic con ex s. Fo
example, a mul i-a med bandi ein o cemen lea ning model
applied in ITSM has demons a ed how AI can dynamically
adjus icke classi ica ions based on human agen eedback,
ensu ing ha au oma ed sugges ions emain con ex ually
ele an and con inuously imp o ing o e ime (Li e al.,
2024). HISE’s s uc u ed app oach o esou ce mobiliza ion
ensu es ha ope an esou ces, such as human skills and AI
analy ics, a e con inuously op imized h ough i e a i e co-
c ea ion p ocesses. Thus, as knowledge e ol es, AI dynami-
cally adjus s i s models o imp o e accu acy and adap abili y
o e ime, ensu ing ha human-AI hyb ids can in eg a e el-
e an esou ces e ec i ely o enhanced decision-making.
HISE also p o ides a go e nance amewo k o s uc-
u ing ins i u ional a angemen s in hyb id in elligen se -
ice ecosys ems. In HISS, AI-d i en se ice ecommenda-
ions mus comply wi h se ice-le el ag eemen s (SLAs),
egula o y s anda ds (such as GDPR o use da a p o ec-
ion), and o ganiza ional policies on AI go e nance. The
ins i u ional laye o HISE ensu es ha hyb id in elligence
does no ope a e in a egula o y acuum; ins ead, i aligns
wi h clea go e nance mechanisms ha de ine accoun -
abili y o AI-d i en ecommenda ions, compliance wi h
Elec onic Ma ke s (2025) 35:63 63 Page 20 o 27
egula o y s anda ds, and mechanisms o bias mi iga ion
and explainabili y. Ins i u ional a angemen s also in lu-
ence he adap a ion o AI-based decision-making ools
o e ime, ensu ing ha models e ol e in alignmen wi h
e hical conside a ions, indus y s anda ds, and egula o y
equi emen s. These go e nance mechanisms a e shaped
no only by p ede ined ules bu also by con inuous in e ac-
ion be ween IT manage s, AI sys em designe s, egula o y
bodies, and se ice desk employees, who nego ia e s and-
a ds, e ine bes p ac ices, and espond o eme ging AI-
d i en isks in eal-wo ld en i onmen s. Hence, AI-d i en
se ice ecommenda ions may need o adap dynamically o
comply wi h eme ging cybe secu i y amewo ks, p i acy
egula ions, and e ol ing ITSM s anda ds, ensu ing ha
hyb id in elligence emains e hically and legally aligned.
The alue o he HISE amewo k becomes pa icula ly
e iden when conside ing he long- e m impac o hyb id
in elligence on scalabili y, adap abili y, and esilience in
ITSM. HISS illus a es how a well-s uc u ed hyb id in el-
ligence se ice model enables IT o ganiza ions o educe he
cogni i e load on se ice agen s, imp o e icke esolu ion
accu acy, and ensu e con inui y o knowledge despi e wo k-
o ce changes (Reinha d e al.,2023). By ollowing HISE’s
p inciples, o ganiza ions can adap hei se ice ope a ions
dynamically, scaling AI in ol emen up o down as needed
wi hou comp omising se ice quali y. Mo eo e , by sys em-
a ically in eg a ing hyb id agency, esou ce mobiliza ion,
and ins i u ional go e nance, HISE ensu es ha AI-d i en
se ice enhancemen s a e sus ainable, e hical, and aligned
wi h human expe ise. This s uc u ed app oach posi ions
HISE as a undamen al amewo k o unde s anding,
designing, and managing hyb id in elligen se ice ecosys-
ems, wi h HISS se ing as a conc e e ins an ia ion o i s
p inciples in IT se ice managemen .
HISS exempli ies he ans o ma i e po en ial o HISE
by showcasing how hyb id in elligence enables IT cus ome
suppo o become mo e adap i e, knowledge-d i en, and
ope a ionally esilien . Ra he han ea ing AI as a simple
e iciency ool, HISE e eals how IT se ice o ganiza ions
can s a egically con igu e hyb id in elligence o enhance
decision-making, op imize se ice wo k lows, and main ain
high-quali y use expe iences. As ITSM landscapes con inue
o e ol e, u u e esea ch should explo e how he HISE
amewo k can guide he de elopmen o mo e ad anced
hyb id in elligence models, ensu ing ha AI-d i en se -
ice suppo emains anspa en , explainable, and aligned
wi h human expe ise. By concep ualizing IT se ice man-
agemen as a hyb id in elligen se ice ecosys em, HISE
p o ides a s uc u ed, heo e ically g ounded app oach o
designing AI-augmen ed se ice en i onmen s, o e ing bo h
esea che s and p ac i ione s a bluep in o he nex gene a-
ion o in elligen IT cus ome suppo .
P oposi ions o u u e esea ch
A e p esen ing and illus a ing ou concep ualiza ion o
HISE, we now discuss he implica ions o esea che s and
p ac i ione s aised by ou concep ual esea ch. Fo his
pu pose, we highligh he e ec s o delibe a e con igu a-
ions o human and a i icial agencies in se ice ecosys ems
by in oducing i e p oposi ions as a ounda ion o u u e
in es iga ion.
Ou amewo k in eg a es exis ing heo e ical concep s
and cons uc s o p opose and desc ibe a dis inc se o
concep s and hei ela ionships wi hin HISE. We adop
widely discussed and ag eed-upon concep s ela ed o he
S-D logic, such as ac o s, esou ces, esou ce in eg a ion,
ins i u ional a angemen s, alue co-c ea ion, and se ice
ecosys ems. To dis inguish HISE om ela ed ye concep-
ually di e en se ice (eco-)sys em concep ualiza ions
(e.g., sma se ice sys ems; Be e ungen e al., 2019), ou
concep ual amewo k p ominen ly includes he inco po-
a ion o a i icial agen s as a sub ype o ma e ial agency
(Leona di, 2011). We in eg a e cu en discussions abou
hyb id in elligence om he ield o human-AI collabo-
a ion, emphasizing ha he app op ia e combina ion and
con igu a ion o human and a i icial agencies can enable
supe io ask pe o mance and, hus, alue co-c ea ion
h ough hyb id in elligen se ice.
The applica ion scena ios p esen ed in he p e ious
sec ion indica e he signi ican impac o AI adop ion on
se ice ecosys ems and how hese AI sys ems a ec he
agency con igu a ion by ac o s, hei esou ce in eg a ion
ac i i ies, hei dynamically e ol ing esou ce s uc u es,
and he o e a ching ins i u ional a angemen s. Conse-
quen ly, we pos ula e ha u u e esea ch should u he
explo e he espec i e ole o hese elemen s in HISE,
building on he i e p oposi ions p esen ed below. Ou
p oposi ions o e po en ial a enues o u u e esea ch
o in es iga e he complex in e play be ween human and
a i icial agencies, esou ces, ins i u ional a angemen s,
and he e olu ion o HISE, ul ima ely con ibu ing o a
deepe unde s anding o how o design and manage hese
ecosys ems e ec i ely.
The accele a ed pace o change wi hin HISE p esen s
bo h oppo uni ies o apid inno a ion and challenges
o adap a ion. AI in eg a ion enables mo e e icien and
esponsi e se ice deli e y, pe sonalized expe iences, and
he eme gence o inno a i e se ice o e ings. Howe e , i
also poses challenges o ac o s wi hin HISE o adap o
apid echnological ad ancemen s, e ol ing ins i u ional
a angemen s, and changing cus ome expec a ions. O gan-
iza ions may need o become mo e agile and adop con inu-
ous lea ning app oaches o emain compe i i e wi hin hese
apidly e ol ing se ice ecosys ems.
Elec onic Ma ke s (2025) 35:63 Page 21 o 27 63
P oposi ion 1: Ac o s
P oposi ion 1: In HISE, he con igu a ion o human and
a i icial agencies o human-AI hyb ids is dynamic and
con ex -dependen .
D awing upon he concep ualiza ion o hyb id in elligence
(Delle mann e al., 2019), ou amewo k p o ides a de ailed
iew o he ole and in e play o human and ma e ial agencies
o echnology-enabled alue co-c ea ion in se ice ecosys-
ems (Ba elheime , 2020). While adi ional concep ualiza-
ions in ma ke ing and IS posi ion IT and AI as me e “ ools”
(i.e., ope and esou ces) o humans, we ollow ecen calls
om esea ch on human-AI collabo a ion in IS esea ch
and adjacen IT- ela ed disciplines (e.g., human-compu e
in e ac ion, compu e science) o acknowledge he g owing
impo ance o IT and especially AI echnologies in many
applica ion con ex s, mani es ed in ma e ial agency ha is
inc easingly equal o human agency (Deme is & Lee, 2018).
This p oposi ion emphasizes ha he balance be ween
human and AI agencies o human-AI hyb ids is no ixed,
bu i is con inuously nego ia ed based on he con ex o
he se ice in e ac ion. Fac o s in luencing his con igu a-
ion include ask complexi y, indi idual skills and p e e -
ences, he pe cei ed use ulness o AI unc ions, and ( he
abili y o ecei e) eal- ime eedback du ing se ice deli -
e y. Fo example, in he semi-au onomous d i ing scena io,
he d i e ’s eliance on AI sys ems may a y depending on
d i ing condi ions, pe sonal com o wi h echnology, and
he capabili ies o he AI assis an . Gene ally, le e aging
AI agencies enables mo e e icien and esponsi e se ice
deli e y, pe sonalized expe iences, and he eme gence o
new se ice o e ings. Howe e , i also poses challenges o
ac o s wi hin HISE o adap o apid echnological ad ance-
men s, e ol ing ins i u ional a angemen s, and changing
cus ome expec a ions. Unde s anding he dynamic con igu-
a ion o agencies is c ucial o designing HISE ha suppo
alue co-c ea ion, subjec o u u e beha io al and design-
o ien ed esea ch.
P oposi ion 2: Resou ces
P oposi ion 2: In HISE, human-AI hyb ids and AI ac o s
o en ha e di e en access o esou ces compa ed o
human-only ac o s, leading o new esou ce in eg a ion
pa e ns based on AI’s abili y o p ocess and analyze
la ge da ase s.
F om a esou ce pe spec i e, he ongoing digi aliza ion
o se ice ecosys ems p omo es he decoupling and ( e-)
combina ion o da a. The agency o AI—and he ypes
o asks ha AI sys ems can pe o m—is subjec o he
a ailabili y o da a, which is cons an ly inc easing due o
esou ce lique ac ion and esou ce densi y (Lusch & Namb-
isan, 2015). Se ice ecosys ems enable all ac o s o exploi
he i ually limi less a ailabili y o da a, bu AI ac o s
o human-AI hyb ids can in eg a e, con igu e, and c e-
a e esou ces in ways ha human-only ac o s canno . Fo
ins ance, in he p ecision ag icul u e scena io, a a me u i-
lizing he AI-d i en analysis o soil condi ions and wea he
pa e ns migh gain a signi ican ad an age o e a a me
who elies solely on adi ional a ming based on expe ience,
as he AI sys em’s abili y o p ocess as amoun s o da a
enables mo e p ecise decision-making, op imizing esou ce
consump ion and inc easing yields.
This p oposi ion highligh s he po en ial disp opo ion
be ween esou ce usage o hyb ids and AI sys ems compa ed
o human ac o s wi hin HISE. AI’s abili y o ga he , p ocess,
and lea n om ex ensi e da ase s can p o ide hyb id and
AI ac o s wi h signi ican ad an ages in a eas like ma ke
p edic ion, isk assessmen , o ope a ional e iciency. This
ad an age migh c ea e powe imbalances wi hin ecosys-
ems, po en ially disad an aging ac o s who (can) ely solely
on human capabili ies and can only capi alize on a limi ed
subse o he a ailable esou ces. O e ime, his phenom-
enon may in ensi y, as hyb id and AI ac o s con ibu e addi-
ional esou ces o he ecosys em ha hey can mobilize,
bu human-only ac o s may ind i challenging o use due o
complexi y, opaci y, and lack o access. Add essing hese
powe imbalances is essen ial o he sus ainable long- e m
success o HISE.
P oposi ion 3: In o ma ion asymme ies
P oposi ion 3: In HISE, unce ain y abou whe he an
ac o is a human-AI hyb id, a human ac o , o an AI ac o
esul s in in o ma ion asymme ies in se ice exchange.
Depending on how asks ha e been alloca ed along he
con inuum be ween ully AI-based and ully human-based
alloca ion, esou ce in eg a ion ac i i ies migh be pe -
o med by humans i s , machines i s , o synch onously
(Recke e al., 2023). T anspa ency ega ding agency con-
igu a ions can os e us and acili a e collabo a ion; how-
e e , s a egic ambigui y can o e compe i i e ad an ages
in ce ain si ua ions. Fo example, o ganiza ions migh
delibe a ely obscu e he ex en o AI in ol emen in hei
alue p oposi ions and o e ings o main ain a compe i i e
ad an age. The esul ing ambigui y abou whe he esou ces
a e being in eg a ed wi h a human-only ac o , an AI-only
ac o , o a human-AI hyb id o en leads o dis o ions o
us among ac o s. In heal hca e scena ios, pa ien s may
be hesi an i hey a e unce ain whe he hei ca e is being
managed by humans o AI sys ems. Con e sely, in compe i-
i e indus ies like inancial ading, i ms may p e e o keep
hei use o AI con iden ial.
Elec onic Ma ke s (2025) 35:63 63 Page 22 o 27
This p oposi ion acknowledges ha he lack o cla i y
abou which ac o s u ilize AI and o wha ex en his impac s
in e ac ions in oduces in o ma ion asymme ies wi hin
HISE. En isioning ha hese asymme ies will inc ease and
become mo e complex wi h highe au onomy o AI ac o s
in he u u e, as well as he nes ed s uc u es ha ac o s can
o m, balancing anspa ency and s a egic in e es s is a
c i ical challenge o HISE. Ensu ing app op ia e le els o
anspa ency can build us and p omo e e ec i e alue co-
c ea ion in HISE, while excessi e opaci y may lead o mis-
us , educed coope a ion, and po en ial e hical conce ns.
P oposi ion 4: Ins i u ional a angemen s
P oposi ion 4: In HISE, he de elopmen o highe -o de
ins i u ions and go e nance mechanisms is essen ial o
e ec i ely add ess he unique challenges posed by in e-
g a ing human-AI hyb ids and AI ac o s.
Ins i u ional a angemen s play a decisi e ole in shap-
ing any ecosys em (O likowski & Ba oudi, 1991), including
HISE, by enabling and cons aining alue co-c ea ion ac i i-
ies. Unde s anding hei ole and impac on human and AI
agencies and how hey a e con igu ed is c ucial. Following
Os om (2009, 2011), we a gue ha wo ypes o ins i u ional
a angemen s need o be s udied in he con ex o HISE:
o mal ins i u ional a angemen s and in o mal ins i u ional
a angemen s.
Fo mal, i.e., legally binding ins i u ional a angemen s
include laws and egula ions ha enable and cons ain he
con igu a ion o agency and alue co-c ea ion. Fo exam-
ple, da a p i acy laws de e mine which esou ces (e.g., pe -
sonal da a) can be used by which con igu a ions o agen-
cies o pe o m a pa icula ask. The ecen ly passed EU
AI Ac egula es he use o AI echnologies in di e en
high- isk se ice con ex s, ensu ing e hical and esponsi-
ble usage (Eu opean Union, 2024). In o mal ins i u ional
a angemen s e e o social no ms, cul u al con ex s, and
indus y p ac ices ha a ec ac o s’ willingness o engage
in ce ain esou ce in eg a ion ac i i ies. Cul u al di e -
ences may in luence how com o able indi iduals a e wi h
AI in ol emen in se ice p o ision. Unde s anding hese
nuances is essen ial when designing HISE and con igu ing
agencies o ensu e accep ance and e ec i e collabo a ion
among ac o s.
This p oposi ion emphasizes he need o obus ins i u-
ional amewo ks o guide he de elopmen , managemen ,
and e olu ion o HISE. Exis ing egula ions and e hical
guidelines may need o be adap ed o expanded o conside
he speci ic challenges and po en ial isks associa ed wi h
AI in eg a ion. Issues like algo i hmic bias, da a p i acy,
e hical use o AI, and he possible displacemen o human
wo ke s equi e ca e ul conside a ion and he de elopmen
o app op ia e go e nance mechanisms o ensu e ha HISE
bene i all ac o s. The EU AI Ac exempli ies he eme ging
egula o y e o s o add ess hese challenges (Eu opean
Union, 2024). Howe e , i is s ill unknown how he in o-
duc ion o such egula o y ins umen s will a ec he design
and managemen o HISE o e ime.
Mo eo e , we an icipa e u u e adap a ions o ins i u-
ional a angemen s as we obse e a shi in he p edominan
“dogma” o AI use. Cu en ly, humans can independen ly
decide whe he o no o collabo a e wi h an AI (e.g., in
heal hca e), bu his migh change. Fo example, pa ien s
a e becoming awa e o he highe accu acy (a leas on a e -
age) o diagnosis when AI is in ol ed. Hence, we expec in
pa s o see a shi owa ds he manda o y use o AI (i.e.,
con igu ing agencies as human-AI hyb ids) in a leas some
applica ion domains due o shi ing cus ome demands and
egula o y equi emen s.
P oposi ion 5: Ecosys em e olu ion
P oposi ion 5: In HISE, AI’s capabili ies accele a e he
pace o ecosys em e olu ion, c ea ing bo h oppo uni ies
and challenges o how hese ecosys ems de elop o e
ime.
Wi hin se ice ecosys ems, ac o s and hei unde lying
agency con igu a ions can ha e a - eaching impac s on
he success and e olu ion o se ice o e ings (Mele e al.,
2018; Nenonen e al., 2018; Vink e al., 2021). In HISE,
human-AI hyb ids and AI ac o s can be expec ed o adap
mo e apidly o en i onmen al changes, gi en hei abili y
o sense and espond as e han human-only ac o s (Va go
& Lusch, 2011).
This p oposi ion highligh s ha apid adap a ion in HISE
may demand and d i e swi adjus men s in ins i u ional
a angemen s, esou ces, and co-c ea ion p ocesses—lead-
ing o bo h oppo uni ies and challenges conce ning he
e olu ion o HISE. Conside ing he eme gence in se ice
ecosys ems (Va go e al., 2023), he AI-d i en adap a ion
o HISE can be iewed as a mul i-laye p ocess in which
AI ac o s, human-AI hyb ids, and human ac o s ecu si ely
shape ins i u ions, esou ces, and co-c ea ion p ocesses.
The in e play be ween eme gence and ins i u ionaliza ion
sugges s ha AI-d i en inno a ions ypically appea i s
as no el ou comes, become embedded as ecu ing pa e ns
in se ice exchange, and e en ually become ins i u ional-
ized h ough egula o y amewo ks (Va go e al., 2023).
This accele a ed e olu ion adds complexi y o human-led
p ocesses and equi es se ice ecosys ems o con inually
adjus o new de elopmen s in he con igu a ion o human
and a i icial agency.
On he one hand, hyb id and AI ac o s acili a e he apid
ollou o inc emen al changes. Se ice ecosys ems, in u n,
Elec onic Ma ke s (2025) 35:63 Page 23 o 27 63
can be comp ehensi ely and sus ainably ( e-)designed in
agile, i e a i e ways hanks o lexible ask alloca ions
be ween human and AI agencies. On he o he hand, his
speed and complexi y in oduce signi ican unce ain y
abou u u e s a es. Va go e al. (2023) emphasize ha such
unce ain y s ems om in ensi ied, ecu si e in e ac ions
AI ac o s in oduce as hey engage wi h human ac o s and
ins i u ions—yielding bo h in ended and unin ended con-
sequences. The accele a ed pace may su pass human and
ins i u ional capaci ies o adap , po en ially causing sys emic
isks. To manage hese isks, delibe a e s a egies a e neces-
sa y o balance con inuous inno a ion wi h ecosys em s a-
bili y, ensu ing ha o e all e olu ion emains aligned wi h
he indi idual and collec i e in e es s o he ac o s in ol ed.
In he semi-au onomous d i ing applica ion scena io,
ADAS p og essi ely ake on asks like adap i e c uise con ol
and collision a oidance, shi ing decision-making au ho i y
om human d i e s o AI. As a esul , ins i u ional a ange-
men s (e.g., oad sa e y laws, liabili y amewo ks) mus e ol e
acco dingly. These AI-d i en ad ancemen s also eshape
esou ce con igu a ion by in eg a ing senso da a, eal- ime
na iga ion based on da a exchange be ween ehicles, and
p edic i e analy ics o ehicle con ol. In u n, hese changes
illus a e a dynamic eedback loop in which egula o y meas-
u es and echnological p og ess con inually in luence each
o he — ein o cing o cons aining possible u u e e olu ion
pa hs (Va go e al., 2023). As AI becomes mo e cen al, se -
ice ecosys ems emain in cons an lux, wi h e ol ing co-c e-
a ion p ocesses and egula ions bo h p opelled by and shaping
AI’s g owing agency.
Conclusion, limi a ions, andou look
This pape p o ides a concep ualiza ion o HISE, a dynamic
ne wo k o in e connec ed human-AI hyb ids, human-only,
and AI-only ac o s, dynamically e ol ing esou ces, and
o e a ching ins i u ional a angemen s ha collec i ely
enable alue co-c ea ion h ough he de elopmen , deploy-
men , ope a ion, and adap a ion o hyb id in elligen se -
ice. Hyb id in elligen se ice e e s o alue co-c ea ion
p ocesses ha in ol e a leas one human-AI hyb id, po en-
ially esul ing in supe io mu ual alue-in-use by alloca ing
asks o human and AI agencies. We p opose ha ou HISE
amewo k p o ides a heo e ically sound basis o be e
unde s anding, designing, and managing hese complex se -
ice ecosys ems. Fo example, i may help o unde s and ha
he ask alloca ion o hyb id ac o s (e.g., indi iduals, eams,
o ganiza ions) is o en no anspa en o o he ac o s, which
can esul in a lack o us among ac o s and migh inally
lead o un ealized alue po en ials wi hin HISE.
We con ibu e o he academic body o knowledge by
in eg a ing p e iously dis inc s eams o esea ch on
(1) hyb id in elligence and human-AI collabo a ion, (2)
S-D logic and se ice ecosys ems, and (3) socio-ma e ial
agency. We p o ide an in eg a ed pe spec i e o unde -
s anding he g owing gene a i e po en ials and mecha-
nisms o AI sys ems in ela ion o exis ing social and
ins i u ional a angemen s a mul iple le els—indi iduals,
eams, o ganiza ions, and socie ies—which is cen al o
he IS discipline. Addi ionally, we p esen i e p oposi-
ions ha call o high-impac u u e esea ch on HISE,
o e ing po en ial a enues o u he in es iga ion.
Beyond hese heo e ical implica ions, ou concep ual
insigh s p o ide guidance o p ac i ione s and policymak-
e s, who can employ he HISE amewo k o analyze and
unde s and he inc easing impo ance o AI in se ice eco-
sys ems and o e lec on how a i icial agencies may a ec
esou ce in eg a ion ac i i ies and ins i u ional a ange-
men s. This implies ha manage s need o moni o hei
own and o he ac o s’ p og ess in adop ing AI sys ems
and ecognize new oppo uni ies o econ igu e human
and a i icial agencies o supe io alue co-c ea ion. Tha
is, ac o s wi hin se ice ecosys ems should delibe a ely
seek new pa e ns o esou ce in eg a ion acili a ed by
he ad ancing capabili ies and use o AI sys ems, whe he
wi hin hei own o ganiza ions o h ough access o o he
ac o s’ AI- ela ed esou ces a he ecosys em le el. A he
mic o le el, he amewo k can guide manage s in e ec-
i ely alloca ing asks o human and a i icial agencies.
This s udy is na u ally subjec o limi a ions ha p o-
ide impe us o u u e esea ch a he same ime. Gi en he
scope and concep ual na u e o his wo k, ou esea ch ep-
esen s only a i s s ep owa d explo ing he consequences
o he deep in eg a ion o human-AI collabo a ion in alue
co-c ea ion p ocesses om a se ice pe spec i e. While he
selec ed scena ios p o ide aluable insigh s in o he appli-
cabili y o he HISE amewo k, o he ele an domains
we e no included. This limi a ion s ems om ou ocus on
indus ies whe e human-AI collabo a ion is mo e elabo a ed
al eady. Fu u e esea ch should in es iga e o he p omising
a eas o u he b oaden and alida e he HISE amewo k’s
applicabili y in di e se se ice ecosys ems.
In addi ion o he concep ual amewo k, u u e design-o i-
en ed, quali a i e, and quan i a i e empi ical esea ch should
de elop and s udy di e en scena ios (e.g., hose p esen ed
in his pape ) o con ibu e o u he de eloping he HISE
amewo k as an adap a ion and ex ension o he S-D logic
di ec ed a unde s anding and explaining socio- echnical phe-
nomena. Empi ical s udies can build on ou p oposi ions o
in es iga e he p ocesses o con igu ing hyb id agencies in
HISE, as well as he p ocesses by which hese agency con-
igu a ions a e shaped and also shape ins i u ional a ange-
men s, i.e., he e olu ion o HISE in he long e m.
Simila o S-D logic and he se ice ecosys em concep ,
he HISE amewo k mo i a es zooming ou o unde s and
Elec onic Ma ke s (2025) 35:63 63 Page 24 o 27
how a mul i ude o di e se ac o s in e ac on di e en le -
els o complex ecosys ems. This app oach o e s a p om-
ising di ec ion o u u e esea ch on hyb id in elligence
and human-AI collabo a ion by conduc ing mul i-le el
(mic o, meso, mac o) analyses o human-AI in e ac ions.
The HISE amewo k acknowledges ha hyb id agency
con igu a ions can exis a all o hese le els, p o iding a
po en ially ui ul lens o e ealing and explaining mul i-
le el dependencies be ween social and ma e ial agency.
S udying agency con igu a ions in HISE is an essen ial
u u e esea ch pa h o IS schola s. In his ega d, he HISE
amewo k can be employed as a ke nel heo y o u u e
design-o ien ed esea ch. Fo example, esea che s can u i-
lize he HISE amewo k o design AI-enabled heal hca e
pla o ms ha op imally alloca e diagnos ic asks be ween
medical p o essionals and AI sys ems. By applying he
p inciples o agency con igu a ion, esou ce in eg a ion,
and adhe ence o ins i u ional a angemen s ou lined in he
amewo k, designe s can de elop sys ems ha enhance
diagnos ic accu acy and e iciency while espec ing pa ien
p i acy and complying wi h medical egula ions. This
illus a es how he HISE amewo k can in o m he c ea-
ion o AI agen s and se ice pla o ms embedded wi hin
se ice ecosys ems, leading o imp o ed alue co-c ea ion.
We posi ha i is an essen ial ask o IS esea che s o
de elop p esc ip i e design knowledge ha suppo s he
de elopmen o e icien and use ul AI agen s o speci ic
asks and con ex s, conside ing hei embedding in se ice
ecosys ems as desc ibed in he HISE amewo k.
Acknowledgemen s We ex end ou g a i ude o all membe s o he
Special In e es G oup Se ices (SIG SVC) in he Associa ion o In o -
ma ionSys ems (AIS) ha pa icipa ed in he SIG SVC annual mee -
ings a ICIS 2022 and ICIS 2023 o sha ing hei eedback andinsigh s
on ea ly e sions o he p esen ed amewo k.
Funding Open Access unding enabled and o ganized by P ojek
DEAL.
Da a A ailabili y The e we e no da a in ol ed in de eloping his con-
cep ual pape .
Decla a ions
Compe ing in e es s The au ho s decla e no compe ing in e es s.
Open Access This a icle is licensed unde a C ea i e Commons A i-
bu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a-
ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce,
p o ide a link o he C ea i e Commons licence, and indica e i changes
we e made. The images o o he hi d pa y ma e ial in his a icle a e
included in he a icle’s C ea i e Commons licence, unless indica ed
o he wise in a c edi line o he ma e ial. I ma e ial is no included in
he a icle’s C ea i e Commons licence and you in ended use is no
pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will
need o ob ain pe mission di ec ly om he copy igh holde . To iew a
copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
Re e ences
Adam, M., Wessel, M., & Benlian, A. (2021). AI-based cha bo s in
cus ome se ice and hei e ec s on use compliance. Elec-
onic Ma ke s, 31(2), 427–445. h ps:// doi. o g/ 10. 1007/
s12525- 020- 00414-7
Akinola, M., Ma in, A. E., & Phillips, K. W. (2018). To delega e
o no o delega e: Gende di e ences in a ec i e associa ions
and beha io al esponses o delega ion. Academy o Manage-
men Jou nal, 61(4), 1467–1491. h ps:// doi. o g/ 10. 5465/ amj.
2016. 0662
An hony, C., Bechky, B. A., & Faya d, A. -L. (2023). “Collabo a ing”
wi h AI: Taking a sys em iew o explo e he u u e o wo k.
O ganiza ion Science, A icle o sc.2022.1651. Ad ance online
publica ion.h ps:// doi. o g/ 10. 1287/ o sc. 2022. 1651
A ie a, A. B., Díaz-Rod íguez, N., Del Se , J., Benne o , A., Tabik, S.,
Ba bado, A., Ga cía, S., Gil-López, S., Molina, D., Benjamins,
R., Cha ila, R., & He e a, F. (2019). Explainable a i icial in el-
ligence (XAI): Concep s, axonomies, oppo uni ies and chal-
lenges owa d esponsible AI.h ps:// doi. o g/ 10. 48550/ a Xi .
1910. 10045
Bai d, A., & Ma uping, L. M. (2021). The nex gene a ion o esea ch
on IS use: A heo e ical amewo k o delega ion o and om
agen ic IS a i ac s. MIS Qua e ly, 45(1), 315–341. h ps:// doi.
o g/ 10. 25300/ MISQ/ 2021/ 15882
Bandu a, A. (2006). Towa d a psychology o human agency. Pe spec-
i es on Psychological Science : A Jou nal o he Associa ion o
Psychological Science, 1(2), 164–180. h ps:// doi. o g/ 10. 1111/j.
1745- 6916. 2006. 00011.x
Ba ile, S., Lusch, R., Reynoso, J., Sa iano, M., & Spoh e , J. (2016).
Sys ems, ne wo ks, and ecosys ems in se ice esea ch. Jou nal
o Se ice Managemen , 27(4), 652–674. h ps:// doi. o g/ 10. 1108/
JOSM- 09- 2015- 0268
Ba elheime , C. (2020). Concep ualizing ask- echnology i o
echnology-pe aded alue co-c ea ion. P oceedings o he 28 h
Eu opean con e ence on in o ma ion sys ems.h ps:// aisel. aisne .
o g/ ecis2 020_ p/ 49
Be e ungen, D., Mülle , O., Ma zne , M., Mendling, J., & om
B ocke, J. (2019). Concep ualizing sma se ice sys ems.
Elec onic Ma ke s, 29(1), 7–18. h ps:// doi. o g/ 10. 1007/
s12525- 017- 0270-5
B aa, M., & Sahay,. (2004). Ne wo ks o ac ion: Sus ainable heal h
in o ma ion sys ems ac oss de eloping coun ies. MIS Qua e ly,
28(3), 337–362. h ps:// doi. o g/ 10. 2307/ 25148 643
B eidbach, C. F., & B odie, R. J. (2017). Engagemen pla o ms in he
sha ing economy. Jou nal o Se ice Theo y and P ac ice, 27(4),
761–777. h ps:// doi. o g/ 10. 1108/ JSTP- 04- 2016- 0071
B enne , W., Za nekow, R., spsampsps Wi ig, H. (1998). In elligen
so wa e agen s: Founda ions and applica ions. Sp inge Be lin
Heidelbe g. h ps:// doi. o g/ 10. 1007/ 978-3- 642- 80484-7
B ozo ić, D., & T egua, M. (2022). The e olu ion o se ice sys ems o
se ice ecosys ems: A li e a u e e iew. In e na ional Jou nal o
Managemen Re iews, 24(4), 459–479. h ps:// doi. o g/ 10. 1111/
ijm . 12287
B ynjol sson, E., & McA ee, A. (2014). The second machine age:
Wo k, p og ess, and p ospe i y in a ime o b illian echnolo-
gies. WW No on & Company.
Caliskan, A., B yson, J. J., & Na ayanan, A. (2017). Seman ics de i ed
au oma ically om language co po a con ain human-like biases.
Science, 356(6334), 183–186. h ps:// doi. o g/ 10. 1126/ scien ce.
aal42 30
Cecez-Kecmano ic, D., Gallie s, R. D., Hen idsson, O., Newell, S.,
& Vidgen, R. (2014). The socioma e iali y o in o ma ion sys-
ems: Cu en s a us, u u e di ec ions. MIS Qua e ly, 38(3),
809–830.h ps:// doi. o g/ 10. 25300/ MISQ/ 2014/ 38:3.3
