January 14, 2026
By Xari Jalil
LAHORE
False and misleading information dominated Pakistan’s digital landscape over the past two years, with artificial intelligence increasingly weaponised to manipulate public opinion, undermine political processes and inflame social divisions, according to a new report by iVerify Pakistan
The AI, Misinformation and Fact-Checking Report 2024–25, released this week, documents how misinformation accelerated during politically volatile moments — particularly the 2024 general elections and a series of regional conflicts in 2025 — while advances in generative AI made false content harder to detect and faster to spread.
iVerify Pakistan, a fact-checking initiative supported by the United Nations Development Programme (UNDP) and implemented by the Centre for Excellence in Journalism (CEJ) at the Institute of Business Administration (IBA), monitored 1,026 potentially false or misleading claims between December 2023 and November 2025. Of these, 513 were investigated in depth, while the rest were discarded due to lack of verifiable evidence or potential harm in amplifying unproven material.
The findings show that nearly four out of five investigated claims — 78.2 per cent — were false, while another 19.7pc were classified as misleading. Only 1.9pc of claims were found to be true.
Politics as the main battleground
Politics overwhelmingly dominated Pakistan’s misinformation ecosystem, accounting for more than 85pc of all claims reviewed. Election-related narratives, judicial developments, political arrests and regional tensions proved especially fertile ground for falsehoods, the report notes.
Social media — particularly X, formerly Twitter — emerged as the primary vector for spreading misleading information, appearing in more than 90pc of the claims tracked. Facebook, Instagram and TikTok played smaller but still significant roles. Encrypted platforms such as WhatsApp, while known to be influential, remained largely unobservable due to technical limitations.
According to the report, supporters of political parties were among the most active disseminators of misleading narratives. PTI supporters featured in nearly a third of the identified claims, followed by Indian-based accounts and supporters of the PML-N. These groups often amplified half-truths, miscontextualised content or recycled old material to advance partisan agendas.
“Misinformation was not just incidental noise — it was frequently strategic,” the report states, noting coordinated campaigns aimed at discrediting opponents, undermining electoral confidence and portraying dissenting groups as “anti-state”.
AI as a ‘force multiplier’
One of the report’s central findings is the rapid rise of AI-enabled misinformation in 2025. iVerify recorded a sixfold increase in AI-related fact-checks compared to the previous year, reflecting the growing accessibility of tools capable of generating realistic images, videos and voice clones.
AI-generated content was used to fabricate speeches, create deepfake endorsements and manipulate existing visuals of political leaders. In several cases, AI was deployed not to create entirely new material, but to subtly alter authentic footage — a phenomenon the report describes as the “AI-fication of reality”.
This shift, the authors warn, has eroded public trust in visual evidence itself. “When even genuine videos can be dismissed as fake, accountability becomes harder to establish,” the report notes.
Women politicians and journalists were disproportionately targeted through sexualised or reputationally damaging synthetic content, reflecting what the report describes as a broader pattern of gendered disinformation.
Elections and voter confusion
During the run-up to the 2024 general elections, AI was used in relatively limited but concerning ways. The report documents manipulated images, doctored audio clips and misleading claims intended to confuse voters — including fabricated statements suggesting election boycotts or candidate withdrawals.
In one instance, a manipulated image of PTI founder Imran Khan circulated widely after his conviction in the cipher case, fuelling speculation about his appearance in jail. Although public scepticism was noticeable, the image nonetheless spread rapidly before being debunked.
The report notes that while Pakistan did not witness the highly sophisticated AI-driven election interference seen in some other countries, attempts to mislead voters remained persistent and potentially damaging to democratic processes.
Wars, conflicts and the fog of misinformation
The intensity of misinformation spiked sharply during regional conflicts in 2025, including the Pakistan–India standoff in May, the Iran–Israel war in June and border tensions with Afghanistan later in the year.
During these periods, iVerify documented a flood of false claims, many involving AI-generated or doctored visuals. Deepfake videos allegedly showing Pakistani leaders conceding defeat, military officials admitting losses or foreign leaders issuing fabricated warnings circulated widely across social media platforms.
Old footage from unrelated conflicts — including Gaza and earlier Middle East escalations — was frequently recycled and misrepresented as evidence of current attacks. The report notes that such visuals were often shared in low resolution to obscure identifying details.
“These conflicts demonstrated how AI can be weaponised to shape public perception during moments of crisis,” the report states, warning that speed and volume often outweighed accuracy in online discourse.
Hate speech and social cohesion
Beyond factual inaccuracies, iVerify identified hate speech as a persistent and dangerous component of misinformation. Of the fact-checks published, 34 addressed explicit hate speech targeting individuals or communities based on religion, ethnicity, gender or political affiliation.
Sectarian and religious misinformation proved particularly volatile, with emotionally charged falsehoods strategically circulated during periods of unrest. The report warns that such narratives deepen mistrust and pose tangible risks to public safety and social cohesion.
Challenges for fact-checkers
The report also outlines structural challenges facing fact-checking efforts in Pakistan, including the sheer speed of viral misinformation, limited access to official information, repeated circulation of debunked claims and growing reliance on unverified AI chatbots for “instant verification”.
Internet shutdowns and social media disruptions during politically sensitive periods further constrained verification efforts, while increasingly sophisticated AI-generated media strained existing detection tools.
Looking ahead
The report calls for fact-checking to be institutionalised as a long-term public-interest function, supported by sustained funding, stronger media partnerships and improved coordination between government bodies and independent verifiers.
It also urges newsrooms to invest in AI literacy, platforms to strengthen accountability mechanisms, and civil society to expand media literacy initiatives.
“As Pakistan moves into an even more AI-saturated future, credible fact-checking initiatives will be essential to protecting democratic discourse and vulnerable communities,” the report concludes.