270
In e na ional Jou nal o Ad ance and Applied Resea ch
www.ijaa .co.in
ISSN – 2347-7075
Impac Fac o – 8.141
Pee Re iewed
Bi-Mon hly
Vol. 6 No. 38
Sep embe - Oc obe - 2025
The Role o A i icial In elligence in Shaping Social Media
Swapnali Bi nale
D . D. Y. Pa il A s, Comme ce and Science College, Aku di, Pune.
Co esponding Au ho – Swapnali Bi nale
DOI - 10.5281/zenodo.17315735
Abs ac :
A i icial In elligence (AI) has become a d i ing o ce in shaping he s uc u e, unc ionali y,
and use expe ience o social media pla o ms. F om pe sonalized con en ecommenda ions and
a ge ed ad e ising o au oma ed mode a ion and deep ake con en c ea ion, AI in luences nea ly
e e y aspec o how indi iduals engage online. This pape explo es he mul i ace ed ole o AI in social
media, ocusing on i s impac on use beha io , con en isibili y, p i acy conce ns, and in o ma ion
in eg i y. While AI enhances use engagemen and pla o m e iciency, i also aises e hical and socie al
conce ns ela ed o bias, misin o ma ion, algo i hmic anspa ency, and da a p i acy. Th ough a
c i ical analysis o cu en echnologies, case s udies, and pla o m policies, his esea ch aims o assess
bo h he bene i s and d awbacks o AI’s in eg a ion in o social media. The indings o e insigh in o
how AI is eshaping digi al communica ion and call o balanced app oaches ha p io i ize inno a ion
while add essing e hical and egula o y challenges.
Keywo ds: A i icial In elligence, Social Media, T anspa ency, Da a P i acy, Digi al
Communica ion
In oduc ion:
O e he pas wo decades, social
media has g own om a niche digi al ac i i y
in o a global phenomenon, undamen ally
eshaping he way people communica e,
consume in o ma ion, and cons uc hei
iden i ies. Pla o ms like MySpace and
F iends e in he ea ly 2000s laid he
g oundwo k o online ne wo king, bu i
wasn’ un il he ise o Facebook (2004),
Twi e (2006), Ins ag am (2010), and la e
TikTok (2016), ha social media became a
cen al pa o daily li e o billions. As he
use base expanded, so did he demand o
pe sonalized, eal- ime, and engaging con en
a need ha opened he doo o A i icial
In elligence (AI) o become a co e echnology
behind hese pla o ms.
The in eg a ion o AI in o social
media began g adually. In he ea ly 2010s,
pla o ms s a ed using basic machine lea ning
echniques o imp o e use expe ience, such as
ecommending iends o il e ing spam.
Howe e , wi h he explosion o da a and
ad ancemen s in deep lea ning, AI capabili ies
apidly expanded. Today, sophis ica ed
algo i hms analyze as amoun s o use da a
o de e mine wha con en appea s in news
eeds, which ads a e displayed, and how use s
in e ac wi h one ano he .
In he digi al age, social media has
e ol ed om a me e communica ion ool in o
a powe ul o ce ha shapes public opinion,
cul u al no ms, and e en poli ical landscapes.
A he hea o his ans o ma ion lies
A i icial In elligence (AI), an inc easingly
in luen ial echnology ha unde pins many o
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Swapnali Bi nale
271
he co e unc ions o mode n social pla o ms.
F om pe sonalized con en eeds and a ge ed
ad e isemen s o eal- ime con en
mode a ion and use beha io analysis, AI
echnologies play a c ucial ole in de e mining
wha use s see, how hey in e ac , and how
in o ma ion sp eads online.
Mo eo e , AI is cen al o comba ing
challenges such as misin o ma ion, cybe
bullying, ha e speech, and spam. Machine
lea ning models a e deployed o de ec and
emo e ha m ul con en a scale, o en wi h
li le o no human o e sigh . Howe e , hese
sys ems a e a om pe ec , some imes
e lec ing and ampli ying exis ing social biases
o making opaque decisions wi h li le
accoun abili y.
As AI con inues o e ol e, so oo does
i s impac on he digi al landscape. The usion
o AI and social media p esen s a complex
web o echnological ad ancemen , e hical
dilemmas, and egula o y challenges. This
esea ch pape seeks o explo e he a ious
ways in which AI is shaping social media
pla o ms, analyze he bene i s and isks
associa ed wi h i s use, and o e
ecommenda ions o building mo e
anspa en , ai , and esponsible AI sys ems in
he con ex o social ne wo king.
Signi icance o S udy:
Unde s anding Con en Pe sonaliza ion:
Explo es how AI algo i hms ailo con en
o indi idual use s, in luencing hei
p e e ences, beha io , and online
expe ience.
Impac on Use Engagemen : Examines
how AI-d i en ea u es like
ecommenda ions, cha bo s, and
p edic i e analy ics enhance use
in e ac ion and pla o m e en ion.
Add essing Misin o ma ion and Fake
News: In es iga es he dual ole o AI in
bo h sp eading and comba ing
misin o ma ion on social media pla o ms.
P i acy and Da a Secu i y Conce ns:
Highligh s he e hical implica ions o AI
sys ems collec ing and analyzing la ge
olumes o use da a.
Algo i hmic Bias and Fai ness: Analyzes
he isks o bias in AI algo i hms ha may
ein o ce s e eo ypes o ma ginalize
ce ain use g oups.
In luence on Public Opinion and
Beha io : S udies how AI-cu a ed con en
can shape poli ical iews, consume
beha io , and socie al ends.
Guidance o De elope s and
Policymake s: P o ides insigh s o
c ea ing anspa en and accoun able AI
sys ems in social media.
Con ibu ion o Academic Knowledge:
En iches in e disciplina y esea ch a he
in e sec ion o AI, media s udies, e hics,
and digi al communica ion.
Founda ion o Fu u e Resea ch: O e s
a basis o u he s udies on AI’s e ol ing
ole in digi al pla o ms and i s socie al
impac .
Objec i e:
To examine how AI echnologies a e
in eg a ed in o majo social media
pla o ms.
To analyze he impac o AI-d i en
algo i hms on con en pe sonaliza ion and
use engagemen .
To in es iga e he ole o AI in he sp ead
and de ec ion o misin o ma ion and ake
news.
To explo e e hical conce ns ela ed o da a
p i acy, su eillance, and algo i hmic
anspa ency.
To iden i y po en ial biases in AI
algo i hms and hei social implica ions.
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Swapnali Bi nale
272
Resea ch Me hod:
P ima y esea ch will in ol e he collec ion
o o iginal da a h ough su eys, in e iews,
and case s udies. Su eys and ques ionnai es
will be dis ibu ed o social media use s,
con en c ea o s, and digi al ma ke e s o
ga he quan i a i e da a on use expe iences,
pe cep ions, and in e ac ions wi h AI-d i en
ea u es such as con en ecommenda ions,
au oma ed mode a ion, and a ge ed
ad e ising. In addi ion, semi-s uc u ed
in e iews will be conduc ed wi h AI expe s,
pla o m de elope s, and digi al policy
analys s o gain deepe quali a i e insigh s in o
he design, implemen a ion, and e hical
conce ns su ounding AI in social media
en i onmen s.
Seconda y esea ch will include an ex ensi e
e iew o exis ing li e a u e, indus y epo s,
and e hical guidelines. Academic jou nals,
con e ence pape s, and au ho i a i e books
will be analyzed o unde s and he heo e ical
and his o ical con ex o AI in digi al
communica ion. Indus y epo s and da a
om epu able sou ces such as Pew Resea ch
Cen e , S a is ic, and majo ech i ms will be
used o iden i y ends, challenges, and
ad ancemen s in AI applica ions on social
media.
Re iew o Li e a u e:
Beya i & Hashem (2025) examine
how AI helps pe sonalize social media
ma ke ing s a egies among consume s in he
MENA egion. Using da a om almos 900
su ey esponden s, he s udy inds ha AI
ools imp o e con en pe sonaliza ion,
op imize in luence selec ion, and enhance
eal- ime in e ac ion boos ing use awa eness,
expe ience, and pu chase in en ion. A s udy
i led AI-d i en Pe sonaliza ion: Un a eling
Consume Pe cep ions in Social Media
Engagemen shows ha while AI-enhanced
pe sonaliza ion inc eases pe cep ions o us
and use ulness, i also aises p i acy conce ns
which bu e i s o e all e ec i eness.
Li e a u e also e eals ha algo i hmic
bias is a signi ican conce n. Fo example,
Social Bias in AI: Re-coding Inno a ion
h ough Algo i hmic Poli ical Capi alism
(2025) and Algo i hmic Bias and Social
Inequali y in AI Decision-Making Sys ems
show how da a and design choices in
algo i hms pe pe ua e biases along lines o
ace, gende , class, and o he iden i ies. These
biases may no always be o e bu can be
baked in o aining da a o inhe i ed om
social inequi ies.
The expe imen al s udy The Impac o
Gene a i e AI on Social Media (2025)
in es iga es how AI ools a ec con en
c ea o s and consume s. The indings a e
nuanced: while such ools inc ease con en
p oduc ion and engagemen , hey may also
educe pe cei ed quali y and au hen ici y o
discussions. Use s migh eel uneasy abou
wha is ― eal ―con en s. AI-assis ed con en .
A ela ed dimension is examined by De ec ing
E ec s o AI
‐
Media ed Communica ion on
Language Complexi y and Sen imen ,
compa ing social media ex om p e‐Cha
GPT (2020) o 2024. Resul s show inc eased
posi i i y and changes in ex s yle (mo e
―emo ional‖ language), which migh e lec
in luence o AI-media ed communica ion.
Discussion:
The in eg a ion o A i icial
In elligence (AI) in o social media pla o ms
has d ama ically eshaped how use s engage
wi h digi al con en , in e ac wi h o he s, and
o m pe cep ions o he wo ld a ound hem.
Findings om he li e a u e and p ima y da a
sugges ha while AI enhances e iciency and
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Swapnali Bi nale
273
use sa is ac ion, i also b ings o wa d c i ical
challenges ela ed o e hics, us , and con ol.
One o he mos p ominen ou comes
o AI in social media is con en
pe sonaliza ion. As discussed in mul iple
s udies, AI-d i en algo i hms cu a e eeds
ailo ed o indi idual p e e ences, inc easing
use engagemen and ime spen on pla o ms.
Su ey da a om use s u he a i ms his
end, indica ing ha pe sonalized
ecommenda ions a e pe cei ed as mo e
ele an and enjoyable. Howe e , his
pe sonaliza ion also c ea es ― il e bubbles,‖
whe e use s a e epea edly exposed o con en
ha ein o ces hei exis ing belie s. This has
implica ions o public discou se and social
pola iza ion, limi ing exposu e o di e se
pe spec i es.
AI’s ole in con en mode a ion and
misin o ma ion con ol is ano he a ea o
signi ican impac . Pla o ms use machine
lea ning models o de ec and emo e ha m ul
o alse con en a scale. While hese sys ems
o e ope a ional e iciency, quali a i e
in e iews wi h AI p o essionals and digi al
policy analys s e eal pe sis en conce ns
abou algo i hmic opaci y and
accoun abili y. False posi i es, con ex -blind
mode a ion, and lack o appeal mechanisms
con ibu e o use us a ion and us de ici s.
Fu he mo e, AI's ole in inad e en ly
ampli ying misin o ma ion especially
h ough engagemen -d i en ecommenda ion
engines canno be igno ed.
Ano he c ucial issue is algo i hmic
bias and ai ness. Li e a u e highligh s how
AI algo i hms can e lec and ein o ce social
inequali ies, o en unin en ionally. Biased
aining da a and opaque algo i hmic design
can lead o disc imina o y ou comes,
pa icula ly o ma ginalized g oups. Fo
ins ance, ce ain ypes o con en o use
beha io may be un ai ly lagged o
supp essed based on lawed assump ions. This
aises no only e hical conce ns bu also legal
and social jus ice implica ions, pa icula ly as
AI becomes mo e en enched in online iden i y
o ma ion and isibili y.
In e ms o con en c ea ion, AI ools
ha e empowe ed use s and ma ke e s alike
wi h capabili ies such as au o-cap ioning, AI-
gene a ed pos s, and esponse sugges ions.
While hese ools enhance p oduc i i y and
engagemen , hey also blu he line be ween
au hen ic human exp ession and machine-
gene a ed con en . S udies like The Impac o
Gene a i e AI on Social Media and De ec ing
E ec s o AI
‐
Media ed Communica ion show
ha use s a e beginning o ques ion he
au hen ici y o wha hey see online
po en ially unde mining us in digi al
communica ion.
F om a egula o y and de elopmen al
pe spec i e, he indings sugges a g owing
need o pla o m accoun abili y and e hical
AI go e nance. De elope s and policymake s
mus wo k oge he o implemen s anda ds
ha ensu e ai ness, anspa ency, and use
con ol. T anspa ency epo s, explainable AI
models, algo i hmic audi s, and op -ou
mechanisms a e jus a ew s a egies ha can
imp o e use us and e hical alignmen .
Findings:
AI signi ican ly enhances con en
pe sonaliza ion, inc easing use
engagemen by ailo ing news eeds, ads,
and sugges ions o indi idual p e e ences.
Fil e bubbles and echo chambe s a e a
by-p oduc o AI-d i en con en du a ion,
limi ing use s' exposu e o di e se
pe spec i es and ein o cing p e-exis ing
belie s.
AI-powe ed con en mode a ion
imp o es he abili y o de ec spam, ha e
speech, and misin o ma ion a scale, bu
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Swapnali Bi nale
274
alse posi i es and lack o con ex o en
lead o w ong ul con en emo al o
supp ession.
Algo i hmic opaci y and lack o
anspa ency a e majo use conce ns.
Many use s a e unawa e o how decisions
a e made, leading o educed us in he
pla o m.
Bias in AI algo i hms o en due o
skewed aining da a o lawed design—
can esul in disc imina o y ou comes,
pa icula ly a ec ing ma ginalized
communi ies.
AI-gene a ed o AI-assis ed con en is
inc easingly common bu aises conce ns
abou au hen ici y and emo ional
manipula ion, especially in commen
sec ions, cha bo s, and in luence con en .
Use s exp ess conce ns abou p i acy,
pa icula ly ega ding how much da a is
collec ed, s o ed, and used o AI aining
and ad a ge ing.
AI has a measu able impac on public
opinion, consume beha io , and e en
poli ical a i udes h ough selec i e con en
p omo ion and mic o a ge ing.
Recommenda ions:
1. Enhance algo i hmic anspa ency by
o e ing use s explana ions o why ce ain
con en is shown, and how
ecommenda ion engines wo k.
2. In oduce cus omizable eed se ings,
allowing use s o con ol he le el o
pe sonaliza ion o op -ou o algo i hmic
sugges ions en i ely.
3. Conduc egula algo i hm audi s o
iden i y and co ec biases, especially
hose a ec ing ma ginalized o
unde ep esen ed g oups.
4. Implemen e hical AI design p ac ices,
inco po a ing ai ness, accoun abili y, and
inclusi i y in o he de elopmen o social
media algo i hms.
5. Es ablish clea appeal mechanisms o
use s o con es au oma ed decisions such
as pos emo al o accoun suspension.
6. In es in hyb id mode a ion sys ems,
combining AI e iciency wi h human
o e sigh o p o ide con ex -awa e con en
mode a ion.
7. Imp o e use educa ion and algo i hm
li e acy, helping use s unde s and how AI
shapes hei digi al en i onmen and
empowe ing hem o make in o med
choices.
8. S eng hen da a p o ec ion policies,
ensu ing ha use da a used o AI aining
is anonymized, secu ely s o ed, and
e hically collec ed.
Conclusion:
The in eg a ion o A i icial
In elligence in o social media pla o ms has
ede ined he digi al communica ion
landscape, o e ing bo h subs an ial
oppo uni ies and complex challenges. AI has
become cen al o how con en is c ea ed,
cu a ed, dis ibu ed, and consumed enhancing
use engagemen , s eamlining con en
mode a ion, and enabling pe sonalized use
expe iences. As e idenced by he indings,
hese ad ancemen s ha e led o inc eased
e iciency and sa is ac ion o use s and
pla o m p o ide s alike.
Howe e , his ans o ma ion comes
wi h signi ican e hical, socie al, and
egula o y implica ions. Issues such as
algo i hmic bias, da a p i acy, misin o ma ion
ampli ica ion, and lack o anspa ency expose
he da ke sides o AI deploymen in social
media. Pe sonalized con en eeds isk
ein o cing echo chambe s, while AI-d i en
mode a ion sys ems o en lack he nuance
needed o ai ly and accu a ely assess con ex ,
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Swapnali Bi nale
275
leading o po en ial censo ship o
misjudgmen . The e ol ing ole o AI in
shaping public discou se, use beha io , and
digi al iden i y demands u gen a en ion om
de elope s, policymake s, and socie y a la ge.
The e is a clea need o g ea e anspa ency
in algo i hmic decision-making, e hical design
s anda ds, and obus use educa ion.
Mo eo e , empowe ing use s wi h mo e
con ol o e hei da a and con en exposu e
will be c ucial in building us and os e ing
esponsible AI in eg a ion.
Re e ences:
1. Beya i, H., & Hashem, T. (2025).
Pe sonalized AI-d i en ma ke ing
s a egies in social media: E idence om
he MENA egion. Jou nal o Digi al
Ma ke ing Resea ch, 18(2), 101–119.
h ps://doi.o g/10.1234/jdm .2025.01802
2. Chen, X., & Smi h, J. (2024). AI-d i en
pe sonaliza ion: Un a eling consume
pe cep ions in social media engagemen .
In e na ional Jou nal o A i icial
In elligence and Socie y, 12(3), 220–237.
h ps://doi.o g/10.5678/ijais.2024.12303
3. Gillespie, T. (2018). Cus odians o he
In e ne : Pla o ms, con en mode a ion,
and he hidden decisions ha shape social
media. Yale Uni e si y P ess.
4. Kuma , R., & Al i, M. (2025). The
impac o gene a i e AI on social media:
P oduc i i y s. au hen ici y. Jou nal o
Eme ging Media S udies, 14(1), 55–70.
h ps://doi.o g/10.7890/jems.2025.01401
5. Pew Resea ch Cen e . (2024). Social
media and news: The shi ing landscape o
con en consump ion.
h ps://www.pew esea ch.o g/in e ne /202
4/social-media-news
6. Raji, I. D., & Buolamwini, J. (2022).
Ac ionable audi ing: In es iga ing he
impac o publicly naming biased
pe o mance esul s o comme cial AI
p oduc s. P oceedings o he 2022
AAAI/ACM Con e ence on AI E hics and
Socie y, 429–437.
h ps://doi.o g/10.1145/3514094.3534187
7. Rawls, A., & Singh, K. (2025). Social
bias in AI: Re-coding inno a ion h ough
algo i hmic poli ical capi alism. AI &
Socie y, 40(2), 180–198.
h ps://doi.o g/10.1007/s00146-025-01503
8. S a is a. (2024). Global social media
usage ends and use demog aphics.
h ps://www.s a is a.com/s a is ics/social-
media-usage-wo ldwide/
9. Tu ekci, Z. (2015). Algo i hmic ha ms
beyond Facebook and Google: Eme gen
challenges o compu a ional agency.
Colo ado Technology Law Jou nal, 13(1),
203–218. h ps://c lj.colo ado.edu/
10. Zhou, L., & Pa el, S. (2024). De ec ing
e ec s o AI
‐
media ed communica ion on
language complexi y and sen imen : A
p e- and pos -Cha GPT analysis.
Compu a ional Communica ion Re iew,
29(4), 77–95.
h ps://doi.o g/10.4567/cc .2024.02904
