Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [406]
UTILIZATION OF ARTIFICIAL INTELLIGENCE TOOLS AMONG MID-LIFERS
IN THE GOVERNMENT SERVICE
Joseph Isaiah S.T inidad
Kimbe ly D. Teso o
Abbel F. Unla
E o enos L. Ema a
Gaudencio G. Abellanosa
Uni e si y o Sou heas e n Philippines, College o De elopmen Managemen G adua e P og am,
Min al Campus, Da ao Ci y
ABSTRACT
This s udy aimed o de e mine he signi ican di e ence in he u iliza ion o A i icial In elligence ools among
mid-li e s in he go e nmen se ice. The esea che s u ilized non-expe imen al quan i a i e and compa a i e
esea ch design o de e mine signi ican di e ences. I is ancho ed on he Ins i u ions speci ically suppo ing
he objec i e o Ta ge 16.6 unde his goal, de elop e ec i e, accoun able and anspa en ins i u ions a all
le els. The s udy was conduc ed wi h 105 employees om he di e en go e nmen agencies. The esea che s
ound ou ha mid-li e s e ealed mode a e le el as o he u iliza ion o AI ools in e ms o cons uc Pe ils.
Mo eo e , hey mani es ed high le el o he cons uc s P omises and Powe s. Las ly, hey disclosed non-
signi ican di e ence in he le el o u iliza ion o AI ools when hey a e g ouped by gende , wo k, s a us, and
leng h o se ice in e ms o cons uc s Pe ils and P omises. Mid-li e s mani es ed signi ican di e ence in he
u iliza ion o AI ools, pa icula ly on he indica o Pe ils on which mid-li e s handling Pe manen posi ion is
leading. No signi ican di e ence was no ed o he indica o s P omises and Powe .
Keywo ds:
A i icial In elligence, Mid-Li e s, Go e nmen Se ice, Pe ils, P omises, Powe
INTRODUCTION
In he con ex o public adminis a ion, employees’ abili y o unde s and and u ilize AI ools is c i ical o
success ul digi al ans o ma ion. Mid-li e s, ypically aged 35–55, occupy essen ial ope a ional and supe iso y
oles ha di ec ly in luence he adop ion and implemen a ion o eme ging echnologies in go e nmen se ice
(AP-NORC, 2024). Unlike younge digi al na i es, mid-li e s o en balance subs an ial ins i u ional expe ience
wi h he need o adap o apidly e ol ing echnological landscapes. Resea ch shows ha his age g oup may
exhibi dis inc pe cep ions owa d AI— anging om en husiasm and pe cei ed use ulness o conce ns abou
e hical isks, job displacemen , and sys em eliabili y (Wong e al., 2025).
Con empo a y s udies highligh ha employee a i udes owa d AI adop ion can be unde s ood h ough h ee
majo dimensions: Pe ils (pe cei ed isks), : P omises (pe cei ed bene i s), and Powe (pe cei ed capabili y o
AI sys ems), as concep ualized by Shum and Lau (2024). High le els o pe cei ed p omises a e o en associa ed
wi h imp o ed pe o mance, educed wo kloads, and enhanced se ice quali y (Huang & Rus , 2021).
Con e sely, pe cei ed pe ils s em om ea o su eillance, da a b eaches, e hical issues, and job insecu i y
(Ge lich e al., 2023; UNESCO, 2021). Pe cei ed powe ela es o AI’s inc easing abili y o equal o exceed
human pe o mance in ou ine o analy ical asks (Zhang & Lu, 2023).
Wi hin he Philippine go e nmen se ing, in e es in AI in eg a ion has g own, pa icula ly in a eas ela ed o
adminis a i e p ocessing, public employmen se ices, on line ansac ions, and da a-d i en decision-making.
Howe e , empi ical s udies ha di ec ly assess AI u iliza ion among mid-li e go e nmen employees emain
sca ce. Unde s anding hei pe cep ions and u iliza ion pa e ns is essen ial because mid-li e s o en se e as
digi al ansi ion leade s and men o s wi hin hei o ganiza ions.
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [407]
Gi en his con ex , he p esen s udy in es iga es he u iliza ion o AI ools among mid-li e s in go e nmen
se ice. I examines hei pe cei ed p omises, pe ils, and powe o AI and analyzes whe he signi ican
di e ences exis ac oss demog aphic cha ac e is ics such as gende , wo k ype, employmen s a us, and leng h
o se ice. This inqui y o e s c ucial insigh s in o he eadiness o he mid-li e public wo k o ce o adop AI
echnologies, he eby suppo ing he b oade goal o os e ing e ec i e and echnologically adap i e
go e nmen ins i u ions.
OBJECTIVES
This s udy aimed o de e mine he signi ican di e ence in he u iliza ion o A i icial In elligence ools among
mid-li e s in he go e nmen se ice. Speci ically, i aims o: 1) De e mine he demog aphic p o ile o he
esponden s. 2) Le el o u iliza ion o AI ools among mid-li e s; 3) Signi ican di e ence he le el o u iliza ion
o AI ools among mid-li e s when hey a e g ouped by gende , wo k, s a us and leng h o Se ice.
METHODOLOGY
This esea ch cons i u es a compa a i e analysis, a me hodical examina ion ha di ec ly compa es wo o mo e
hings o iden i y hei simila i ies and di e ences (Jean Kaluza, 2023). The p ima y aim o he s udy is o
asce ain signi ican di e ences in he le el o u iliza ion o AI ools among mid-li e s based on hei
demog aphic p o iles. Da a o his s udy we e ga he ed h ough an adap i e ques ionnai e om “Pe ils, powe
and p omises: La en p o ile analysis on he a i udes owa ds a i icial in elligence (AI) among middle-aged and
olde adul s in Hong Kong” (Shum e al., 2024). The s udy in ol ed 105 esponden s om he di e en
go e nmen agencies.
The ques ionnai e comp ised wo dis inc sec ions. The i s sec ion cap u ed demog aphic cha ac e is ics,
including gende , wo k, s a us and leng h o se ice. The second sec ion consis ed o 20 i ems ac oss h ee
indica o s. The ollowing s a is ical ools we e used such as mean, - es , and ANOVA.
RESULTS AND DISCUSSION
This chap e p esen s he esul s o he s udy, including he desc ip i e s a is ics (mean and s anda d de ia ion),
along wi h he co esponding in e p e a ions and analyses. Tables a e u ilized o illus a e he indings, and he
discussion o bo h abula and g aphical da a is p o ided o acili a e cla i y and unde s anding.
Demog aphic P o ile
P esen ed in Table 1 is he demog aphic p o ile o he esponden s. The e we e 105 esponden s, 52 male and 53
emale equi alen 49.5% and 50.5% espec i ely. Mos o he esponden a e male. The e we e 5 esponden s
occupying eaching posi ion equi alen o 4.8 pe cen while non- eaching was 100 equi alen o 95.2 pe cen .
Mos o he esponden s we e non- eaching posi ions. Fu he mo e, ou o 105, 71 o 67.6 pe cen o he
esponden s hold a pe manen s a us while 32.4 pe cen o 34 esponden s equi alen o 74.3 pe cen a e non-
pe manen . Las ly, ega ding he leng h o se ice, 78 esponden s se ed o less han 10 yea s, 19 esponden s
se ed o 11-20 yea s while 8 esponden s se ed o o e 21 yea s. Mos o he esponden s se ed o less
han 10 yea s.
This inding e lec s b oade ends obse ed in public-sec o employmen in ecen yea s. The nea ly equal
dis ibu ion o male and emale esponden s aligns wi h global pa e ns showing ha go e nmen ins i u ions a e
among he mos gende -balanced wo kplaces (OECD, 2023). The p edominance o non- eaching and
adminis a i e pe sonnel in he sample is consis en wi h na ional and in e na ional epo s indica ing ha mos
mid-le el go e nmen unc ions a e occupied by echnical and adminis a i e s a a he han ins uc ional
pe sonnel (AP-NORC, 2024). The high p opo ion o pe manen employees likewise mi o s public-sec o
wo k o ce s uc u es in Sou heas Asia, whe e pe manen appoin men s emain he mos common employmen
s a us in na ional go e nmen agencies (Asian De elopmen Bank, 2021). Addi ionally, he la ge numbe o
esponden s wi h ewe han en yea s in se ice suppo s indings ha many go e nmen s ha e expanded hi ing
e o s in esponse o mode niza ion ini ia i es, digi al ans o ma ion demands, and e ol ing public se ice
needs (Wo ld Bank, 2024). These demog aphic pa e ns co espond wi h s udies no ing ha mid-li e s o en
possess di e se leng hs o se ice as go e nmen agencies unde go ansi ions owa d mo e echnology-enabled
sys ems (Duck & E nes 2025).
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [408]
Indica o
F equency
Pe cen
Gende
Male
52
49.5
Female
53
50.5
To al
105
100
Type o Wo k
Teaching
5
4.8
Non-Teaching
100
95.2
To al
105
100
S a us
Pe manen
71
67.6
Non-Pe manen
34
32.4
To al
105
100
Leng h o Se ice
1-10
78
74.3
11-20
19
18.1
21 abo e
8
7.6
To al
105
100
Table 1. Demog aphic P o ile
U iliza ion o AI Tools Among Mid-li e s
Pe ils. P esen ed in Table 2 is he le el o AI ools among mid-li e s. Fo he indica o pe ils, he esponden s
e ealed mode a e le el o he means sco e o 3.40. The mid-li e s claimed ha a i icial in elligence is used o
spy on people wi h he mean sco e 3.66 o high le el. The mid-li e s claimed ha hey some imes shi e wi h
discom o when hey hink abou u u e uses o A i icial In elligence wi h mean sco e o 3.19 o mode a e
le el. Fu he , da a e ealed ha he mid-li e s claimed ha a i icial in elligence o en akes con ol o people,
h ea ening, use o spy people and ake con ol o people. They some imes claim ha AI poses dange when
hinking abou i s u u e uses. Mo eo e , mid-li e s some imes hink ha hey will su e i AI is u ilized
une hically and ha i may some imes commi equen e o s.
P omises. As shown in he able below, mid-li e s o en pe cei e ha AI is exci ing wi h a mean sco e o 3.97
o high le el o u iliza ion. While hey pe cei e ha o ou ine ansac ions, hey would a he in e ac wi h an
a i icially in elligen sys em han wi h a human. The mid-li e s ind he idea o AI exci ing because i can be
used in hei cu en posi ion as i o en p o ides new economic oppo uni ies and p o ides posi i e impac s so
ha he u u e o socie y will bene i om AI u iliza ion. Las ly, o ou ine ansac ions, mid-li e s would a he
in e ac wi h an AI sys em han wi h a human.
Powe . As exhibi ed in he able, he esponden s showed a high le el o app ecia ion o he bene icial
applica ions o AI wi h a high le el o u iliza ion, and mean sco e o 4.04. Addi ionally, mid-li e s o en hink
ha AI has he po en ial o ou pe o m humans in ou ine jobs and con ey hei o e all imp essi eness wi h
wha AI can do.
U iliza ion o AI Tools
SD
Mean
Desc ip i e
Le el
Pe ils
1.02
3.4
Mode a e
I hink A i icial In elligence is dange ous.
1.25
3.41
Mode a e
A i icial In elligence migh ake con ol o people.
1.22
3.58
High
I ind A i icial In elligence h ea ening.
1.24
3.5
High
I shi e wi h discom o when I hink abou u u e uses o A i icial
In elligence.
1.41
3.19
Mode a e
People like me will su e i A i icial In elligence is used mo e and mo e.
1.48
3.28
Mode a e
O ganiza ions use A i icial In elligence une hically.
1.17
3.34
Mode a e
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [409]
A i icial In elligence is used o spy on people.
1.28
3.66
High
I hink a i icially in elligen sys ems make many e o s.
1.11
3.24
Mode a e
P omises
0.78
3.76
High
A i icially in elligen sys ems can help people eel happie .
1
3.74
High
I am in e es ed in using a i icially in elligen sys ems in my daily li e.
1.12
3.72
High
A i icial In elligence is exci ing.
0.98
3.97
High
Much o socie y will bene i om a u u e ull o A i icial In elligence.
0.92
3.86
High
I would like o use A i icial In elligence in my own job.
0.96
3.84
High
A i icial In elligence can p o ide new economic oppo uni ies.
0.95
3.95
High
A i icial In elligence can ha e posi i e impac s on people's wellbeing.
0.98
3.7
High
Fo ou ine ansac ions, I would a he in e ac wi h an a i icially
in elligen sys em han wi h a human.
1.21
3.28
Mode a e
Powe
0.77
3.72
High
The e a e many bene icial applica ions o A i icial In elligence.
0.89
4.04
High
An a i icially in elligen agen would be be e han an employee in many
ou ine jobs.
1.06
3.54
High
A i icial In elligen sys ems can pe o m be e han humans.
1.22
3.56
High
I am imp essed by wha A i icial In elligence can do.
1.1
3.72
High
Table 2. Le el o U iliza ion o AI Tools Among Mid-li e s
The pa e ns obse ed in he esponden s’ pe cep ions o AI, pa icula ly he mode a e conce ns ela ed o pe ils
and he s ong acknowledgmen o p omises and powe a e consis en wi h ecen empi ical esea ch. Se e al
s udies show ha indi iduals, especially mid-ca ee and olde wo ke s, equen ly exp ess cau ion abou AI due
o pe cei ed isks such as su eillance, da a misuse, algo i hmic e o s, and loss o con ol (Ge lich e al., 2023;
UNESCO, 2021). These conce ns mi o he esponden s’ mode a e ag eemen ha AI can be h ea ening, may
commi e o s, and could lead o une hical applica ions i no p ope ly egula ed. A he same ime, li e a u e
om 2020 o 2025 highligh s ha wo ke s also ecognize he subs an ial bene i s o AI, including e iciency
gains, imp o ed public se ice deli e y, and enhanced job pe o mance (Huang & Rus , 2021; Wong e al.,
2025). This suppo s he high mean sco es in he P omises cons uc , whe e mid-li e s iew AI as exci ing,
help ul o ou ine asks, and capable o c ea ing economic and o ganiza ional alue. Fu he mo e, he s ong
ag eemen unde he Powe cons uc aligns wi h global indings demons a ing ha employees inc easingly
belie e AI can ou pe o m humans in ou ine and epe i i e unc ions due o i s analy ical and compu a ional
s eng hs (Zhang & Lu, 2023). O e all, hese s udies alida e he esponden s’ mixed bu gene ally posi i e
a i udes owa d AI, e lec ing bo h cau ious op imism and ecogni ion o AI’s expanding ole in go e nmen
wo k.
U iliza ion o AI Tools when G ouped by Gende
Pe ils. P esen ed in Table 3 is he non-signi ican di e ence in he le el o u iliza ion o AI ools among mid-
li e s when hey a e g ouped in gende . The mid-li e s mani es ed non-signi ican di e ence as he indica o
pe ils e lec ed in he - alue o 0.79 wi h he p- alue o 0.43 which is less han 0.05 le el o signi icance. The
esul is no signi ican and he accep ance o he null hypo hesis. This implies ha male and emale mani es ed
equal pe cep ion as o he u iliza ion o AI ools along his indica o .
P omises. As shown in he able, mid-li e s disclosed non-signi ican di e ences on he cons uc p omise
which hen showed a - alue o 1.64 wi h he p- alue o 0.10 which is less han 0.05 le el o signi icance. The
esul is no signi ican hence he null hypo hesis is accep ed. This in e s ha male and emale mid-li e s,
showed equal le els o pe cep ion owa ds he u iliza ion o AI o he indica o o p omises.
Powe . Indica ed in he able, mid-li e s con eyed non-signi ican di e ence on he cons uc powe which
showed a - alue o 0.69 wi h he p- alue o 0.49 which is lesse han 0.05 le el o signi icance. The esul is no
signi ican hus he null hypo hesis is accep ed. This indica es ha male and emale mid-li e s, showed equal
le els o pe cep ion owa ds he u iliza ion o AI o he indica o o powe .
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [410]
U iliza ion o AI Tools
Male
Female
- alue
p- alue
Decision on H0
Pe ils
3.48
3.32
0.79
0.43
Accep
P omises
3.88
3.63
1.64
0.10
Accep
Powe
3.77
3.67
0.69
0.49
Accep
Table 3. Signi ican Di e ence in he Le el o U iliza ion o AI Tools when G ouped by Gende
No signi ican di e ences we e ound ac oss all cons uc s, sugges ing simila AI pe cep ions among male and
emale mid-li e s consis en wi h global indings (Zou & Schiebinge , 2022).
U iliza ion o AI Tools when G ouped by Wo k
Pe ils. P esen ed in Table 4 is he non-signi ican di e ence in he le el o u iliza ion o AI ools among mid-
li e s when hey a e g ouped by wo k. The mid-li e s mani es ed non-signi ican di e ence as he indica o
pe ils e lec ed in he - alue o 0.34 wi h he p- alue o 0.70 which is less han 0.05 le el o signi icance. The
esul is no signi ican and he accep ance o he null hypo hesis. This implies ha male and emale mani es ed
equal pe cep ion as o he u iliza ion o AI ools along his indica o .
P omises. As shown in he able, mid-li e s disclosed non-signi ican di e ences on he cons uc p omise
which showed a - alue o 1.61 wi h he p- alue o 0.11 which is lesse han 0.05 le el o signi icance. The
esul is no signi ican hence he null hypo hesis is accep ed. This implies ha male and emale mid-li e s,
showed equal le els o pe cep ion owa ds he u iliza ion o AI o he indica o o p omises.
Powe . Las ly, mid-li e s con eyed non-signi ican di e ences on he cons uc powe which hen showed a -
alue o -1.59 wi h he p- alue o 0.12 which is lesse han 0.05 le el o signi icance. The esul is no
signi ican ; he e o e, he null hypo hesis is accep ed. This implies ha male and emale mid-li e s, showed
equal le els o pe cep ion owa ds he u iliza ion o AI o he indica o o powe .
U iliza ion o AI Tools
Teaching
Non-Teaching
- alue
p- alue
Decision on H0
Pe ils
3.58
3.39
0.34
0.70
Accep
P omises
4.30
3.73
1.61
0.11
Accep
Powe
4.25
3.69
1.59
0.12
Accep
Table 4. Signi ican Di e ence in he Le el o U iliza ion o AI Tools when G ouped by Wo k
U iliza ion o AI Tools when G ouped by S a us
Pe ils. Disclosed in Table 5 is he signi ican di e ence in he le el o u iliza ion o AI ools among mid-li e s
when hey a e g ouped by s a us as shown in he - alue o 3.07 wi h a p- alue o .000 which is lesse han 0.05
le el o signi icance. The esul is signi ican and he ejec ion o he null hypo hesis. This implies ha mid-
li e s ha ing pe manen s a us showed highe le el o u iliza ion o AI ools compa ed o non-pe manen s a us.
The indings suppo he s udy o (Duck & E nes 2025) as AI ools usage enhances employees’ wo k
pe o mance in key a eas and i educed bu nou and lowe ed u no e a es.
P omises. As shown in he able, mid-li e s disclosed non-signi ican di e ences on he cons uc p omise
which showed a - alue o 1.14 wi h he p- alue o 0.26 which is less han 0.05 le el o signi icance. The esul
is no signi ican hence he null hypo hesis is accep ed. This implies ha mid-li e s ha ing pe manen and non-
pe manen s a us, displayed equal le el o pe cep ion owa ds he u iliza ion o AI along wi h his indica o .
Powe . Mid-li e s con eyed non-signi ican di e ence in he cons uc powe which hen showed a - alue o
1.94 wi h he p- alue o 0.06 which is lesse han 0.05 le el o signi icance. The esul is no signi ican
he e o e he null hypo hesis is accep ed. This implies ha mid-li e s occupying pe manen and non-pe manen
posi ions showed equal le el o pe cep ion owa ds he u iliza ion o AI o his indica o .
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [411]
U iliza ion o AI Tools
Pe manen
Non-Pe manen
- alue
p- alue
Decision on H0
Pe ils
3.61
2.98
3.07
0.00
Rejec
P omises
3.82
0.82
1.14
0.26
Accep
Powe
3.82
0.83
1.94
0.06
Accep
Table 5. Signi ican Di e ence in he Le el o U iliza ion o AI Tools when G ouped by S a us
U iliza ion o AI Tools when G ouped by Leng h
Pe ils. P esen ed in Table 6 is he non-signi ican di e ence in he le el o u iliza ion o AI ools among mid-
li e s as classi ied by wo k. The mid-li e s mani es ed non-signi ican di e ence as he indica o pe ils e lec ed
in he - alue o 1.85 wi h he p- alue o 0.162 which is lesse han 0.05 le el o signi icance. The esul is no
signi ican and he accep ance o he null hypo hesis. This implies ha male and emale mani es ed equal
pe cep ion as o he u iliza ion o AI ools along his indica o o pe ils.
P omises. Fu he mo e, mid-li e s disclosed non-signi ican di e ences on he cons uc p omise which hen
showed a - alue o 0.524 wi h he p- alue o 0.594 which is lesse han 0.05 le el o signi icance. The esul is
no signi ican hence he null hypo hesis is accep ed. This implies ha male and emale mid-li e s, showed equal
le els o pe cep ion owa ds he u iliza ion o AI o he indica o o p omises.
Powe . Finally, mid-li e s con eyed non-signi ican di e ence on he cons uc powe which hen showed a -
alue o 1.34 wi h he p- alue o 0.27 which is lesse han 0.05 le el o signi icance. The esul is no signi ican
hence he null hypo hesis is accep ed. This implies ha male and emale mid-li e s, showed equal le els o
pe cep ion owa ds he u iliza ion o AI o he indica o o powe .
U iliza ion o AI Tools
1-10
11-20
21-abo e
F- alue
p- alue
Decision on
H0
Pe ils
3.30
3.80
3.46
1.85
0.162
Accep
P omises
3.72
3.83
3.99
0.524
0.594
Accep
Powe
3.64
3.92
3.94
1.34
0.267
Accep
Table 6. Signi ican Di e ence in he Le el o U iliza ion o AI Tools when G ouped by Leng h o Se ice
CONCLUSION
O e all, mid-li e s in go e nmen se ice demons a e a gene ally posi i e pe cep ion o AI ools, especially in
e ms o hei use ulness, e iciency, and capaci y o enhance o ganiza ional pe o mance. This aligns wi h
global indings ha public-sec o employees ecognize AI’s po en ial o s eamline adminis a i e p ocesses and
imp o e decision-making (Wo ld Bank, 2024; Wong e al., 2025). Al hough conce ns abou isks such as
p i acy issues, algo i hmic e o s, and e hical misuse emain mode a e, hese app ehensions a e common among
mid-ca ee wo ke s who a e adap ing o apid echnological changes (Ge lich e al., 2023). The minimal
di e ences ac oss demog aphic ca ego ies sugges ha a i udes owa d AI among mid-li e s a e ela i ely
consis en , indica ing a gene ally ecep i e wo k o ce capable o suppo ing digi al go e nance ini ia i es.
Howe e , he signi ican a ia ion obse ed among pe manen employees implies ha hose wi h long- e m
oles may eel mo e accoun able o po en ial consequences o AI misuse and hus exp ess heigh ened cau ion.
S udies show ha pe manen s a o en pe cei e g ea e ins i u ional esponsibili y and he e o e equi e mo e
s uc u ed guidance in na iga ing eme ging echnologies (Duck & E nes 2025). These indings unde sco e he
need o con inuous AI- ela ed capaci y-building p og ams ha emphasize no only echnical skills bu also
esponsible, e hical, and anspa en use o AI ools, as ecommended by (UNESCO, 2021) and (OECD, 2023).
By add essing bo h con idence and conce ns, go e nmen agencies can be e p epa e hei mid-li e wo k o ce
o engage wi h AI inno a ions e ec i ely and esponsibly.
RECOMMENDATION
Based on he indings o he s udy and suppo ed by ecen esea ch on public-sec o AI adop ion, he ollowing
ecommenda ions a e p oposed:
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [412]
1. S eng hen AI Li e acy and Skills T aining o Go e nmen Mid-Li e s.
Gi en ha mid-li e s demons a ed s ong in e es in he bene i s and capabili ies o AI, go e nmen agencies
should ins i u ionalize egula capaci y-building p og ams ha ocus on p ac ical AI usage, da a li e acy, and
in eg a ion in o adminis a i e wo k lows. S udies show ha AI adop ion is signi ican ly highe when
employees ecei e s uc u ed aining and o ganiza ional suppo (Wong e al., 2025; Wo ld Bank, 2024).
T aining should include hands-on sessions in ol ing ools al eady used in public o ices, such as au oma ed ex
gene a o s, da a analy ics pla o ms, and AI-d i en public se ice applica ions.
2. Implemen E hical and Responsible AI Use Wo kshops.
Because mode a e conce ns abou AI- ela ed isks we e obse ed, i is ecommended ha agencies inco po a e
e hics- ocused aining ha add esses algo i hmic bias, da a p i acy, anspa ency, and esponsible use o
au oma ed sys ems. UNESCO (2021) emphasizes ha public-sec o wo ke s mus be ained o unde s and he
e hical implica ions o AI, while OECD (2023) s esses he impo ance o e hical go e nance amewo ks.
Wo kshops should include eal-wo ld case examples o help employees make in o med decisions when using AI
ools.
3. P o ide Ta ge ed Suppo o Pe manen Employees Showing Highe Pe cei ed Risks.
Pe manen employees exhibi ed signi ican ly highe conce ns unde he “Pe ils” cons uc . This indica es a need
o a ge ed in e en ions ha add ess hei ele a ed sense o esponsibili y and accoun abili y. Acco ding o
(Duck & E nes 2025), employees wi h longe enu e o pe manen oles end o be mo e cau ious owa d digi al
inno a ions due o conce ns abou long- e m job implica ions and ins i u ional isks. Tailo ed men o ing,
coaching, and open dialogues on AI go e nance can help educe app ehensions.
4. Ins i u ionalize Clea AI Go e nance Policies Ac oss Go e nmen Agencies.
To add ess conce ns abou su eillance, misuse, and algo i hmic e o s, agencies should de elop o adop
s anda dized AI go e nance guidelines. These may include p o ocols on da a p i acy, anspa ency,
accoun abili y, and human o e sigh . Global amewo ks such as UNESCO's Recommenda ion on he E hics o
A i icial In elligence (2021) and OECD’s AI P inciples (2023) can se e as empla es o local policy
o mula ion.
5. P omo e a Cul u e o Human - AI Collabo a ion.
Since mid-li e s al eady pe cei e AI as use ul and capable, agencies should encou age wo k models ha
highligh AI as a suppo i e ool a he han a eplacemen o human wo ke s. Resea ch shows ha wo ke s
adap mo e posi i ely when AI is amed as a collabo a i e sys em ha enhances human decision-making a he
han eplacing oles (Ge lich e al., 2023; Huang & Rus , 2021). Leade ship communica ion should ein o ce
his pe spec i e o build us and con idence in AI echnologies.
6. Conduc Con inuous Moni o ing and E alua ion o AI U iliza ion in Go e nmen Se ice.
To ensu e e ec i e adop ion, agencies should egula ly assess how AI ools a e used, iden i y challenges, and
ga he eedback om employees. Con inuous e alua ion helps align AI ini ia i es wi h o ganiza ional goals and
wo k o ce needs, as ecommended by he Wo ld Bank (2024) and AP-NORC (2024). This moni o ing can guide
u u e policy adjus men s and aining p og ams.
ACKNOWLEDGEMENT
We would like o ex end ou since e g a i ude o D . Gaudencio G. Abellanosa, o his in aluable assis ance and
guidance h oughou ou esea ch endea o . His expe ise and suppo ha e been ins umen al in shaping he
di ec ion and ou comes o ou s udy. We deeply app ecia e his dedica ion and commi men o helping us
na iga e h ough he in icacies o ou esea ch p ocess.
REFERENCES
[1] AP-NORC Cen e o Public A ai s Resea ch. (2024). Wo kplace echnology and a i udes o mid-ca ee
p o essionals. AP-NORC.
Volume-09 Issue 11, No embe -2025 ISSN: 2456-9348
Impac Fac o : 8.232
In e na ional Jou nal o Enginee ing Technology Resea ch & Managemen
(IJETRM)
h ps://ije m.com/
IJETRM (h p://ije m.com/) [413]
[2] Asian De elopmen Bank. (2021). Public sec o managemen epo : Wo k o ce s uc u es in Sou heas
Asia. ADB Publica ions.
[3] Duck, A. S., & Ea nes , D. (2025). Impac o A i icial In elligence on Employee Job Pe o mance,
Communica ion, and Job Secu i y.
[4] Ge lich, R., Baumann, A., & Hunning, T. (2023). AI anxie y in he wo kplace: An eceden s and
consequences. Jou nal o Business Resea ch, 161, 113897. h ps://doi.o g/10.1016/j.jbus es.2022.113897
[5] Huang, M.-H., & Rus , R. (2021). A s a egic amewo k o a i icial in elligence in ma ke ing. Jou nal o
he Academy o Ma ke ing Science, 49, 30–50. h ps://doi.o g/10.1007/s11747-020-00730-0
[6] OECD. (2023). Public employmen ou look 2023. OECD Publishing.
[7] Shum, A., & Lau, J. (2024). Pe ils, powe , and p omises: La en p o ile analysis on a i udes owa d
a i icial in elligence among middle-aged and olde adul s. Compu e s in Human Beha io , 152, 107154.
h ps://doi.o g/10.1016/j.chb.2023.107154
[8] UNESCO. (2021). Recommenda ion on he e hics o a i icial in elligence. UNESCO Publishing.
h ps://unesdoc.unesco.o g/
[9] Wong, M., Chiu, S., & Ng, E. (2025). De e minan s o AI adop ion in en ion among go e nmen wo ke s
in Asia. Asia Paci ic Jou nal o Public Adminis a ion, 47(1), 1–19.
[10] Wo ld Bank. (2024). Go e nmen AI eadiness epo 2024. Wo ld Bank Publica ions.
h ps://www.wo ldbank.o g/
[11] Zhang, T., & Lu, Y. (2023). Pe cei ed bene i s and isks o AI-enabled se ices: A sys ema ic e iew.
In o ma ion Sys ems F on ie s, 25, 765–780. h ps://doi.o g/10.1007/s10796-022-10343-9
[12] Zou, J., & Schiebinge , L. (2022). Ensu ing gende - ai AI o public se ice wo kplaces. Na u e Machine
In elligence, 4, 986–995.
