Lai Mohammed: AI deepfakes threaten journalism, democratic accountability

Politics Spice Special Features Education Sex & Relationship Interview Columns Opinion Advertise with us Lai Mohammed: AI deepfakes threaten journalism, democratic accountability News November 12, 2026 12:05 am Lai Mohammed

Former Minister of Information and Culture, Lai Mohammed, has warned that artificial intelligence could make it increasingly difficult to hold public figures accountable by enabling them to dismiss genuine photographs, videos and audio recordings as fabricated.

He said the growing sophistication of AI-generated content was creating a dangerous situation in which the technology could not only manufacture convincing false evidence but also provide a convenient defence against authentic evidence, deepening public distrust in journalism and democratic institutions.

Mohammed spoke on Wednesday at Caleb University’s maiden public lecture in Imota, Lagos, titled Fake News, Disinformation and Artificial Intelligence: Defending Truth and Public Trust in the Digital Age.

Describing the phenomenon as the “liar’s dividend,” a term developed by legal scholars Robert Chesney and Danielle Citron, he said the widespread knowledge that artificial intelligence could generate convincing deepfakes might allow people confronted with genuine evidence to discredit it by claiming it was artificially produced.

“Imagine a politician making a genuine statement that is recorded on video. If the politician later says, ‘That video is AI-generated,’ some members of the public may believe the denial simply because they know that AI can create fake videos,” he said.

He added, “The existence of deepfakes therefore creates what the legal scholars Robert Chesney and Danielle Citron have called the ‘liar’s dividend’: once society knows that convincing evidence can be fabricated, a person confronted with genuine evidence may claim that the evidence itself is fabricated.”

The former minister said the threat extended beyond the production of false content to the possibility of an information environment in which citizens could no longer establish what was authentic.

“The danger, therefore, is not simply that AI can create better lies. The greater danger is that AI may create an environment in which people stop believing anything,” he said.

He warned that society could move into what he described as “information nihilism”, where people lose faith in their ability to distinguish truth from falsehood because of competing information, propaganda, conspiracy theories and misinformation.

“And when citizens no longer believe anything, they become vulnerable to believing anything,” he added.

Mohammed, who served as information minister between 2015 and 2023, said the problem had become more sophisticated since the Federal Government launched its National Campaign Against Fake News and Misinformation in July 2018.

Recalling the campaign, he said the government had recognised that misinformation was no longer merely a communication problem but a threat to national security, social cohesion and democratic stability.

“There is an epidemic sweeping the world. It is the epidemic of Fake News,” he recalled saying at the campaign’s launch in Abuja.

He cited incidents in which false information had triggered violence, damaged personal relationships and caused public panic, including the 2018 mob attacks in India following rumours circulated on WhatsApp that people were abducting children.

He also recounted a false report alleging that he had $1.2bn in his bank account, which he said forced him to explain to his wife that the claim was fabricated.

More recently, he cited an incident in Ogun State in which a TikTok content creator allegedly staged a video of a bandit attack to attract followers, triggering panic in the community.

He said such incidents demonstrated that the dangers of disinformation did not depend on sophisticated technology, as ordinary smartphones and social media could already be used to spread damaging falsehoods.

On the political implications, Mohammed said AI could make it easier to circulate fabricated audio recordings of candidates, manipulated photographs of electoral violence and false government documents, potentially influencing public perceptions and electoral participation.

He recalled that his ministry engaged technology companies, including Google and Meta, ahead of the 2023 general elections over concerns about parody accounts, cloned media platforms, manipulated campaign speeches, fabricated opinion polls and false reports of threats and violence.

He, however, cautioned against using the fight against disinformation to restrict legitimate criticism, insisting that governments must distinguish between deliberate deception and honest error, as well as criticism and incitement.

“A democratic government must not demand that citizens believe everything it says. Its responsibility is to make accurate, timely and verifiable information available and to give citizens the tools to assess competing claims,” he said.

Mohammed also challenged journalists to strengthen verification as AI makes fabricated content increasingly difficult to distinguish from authentic material.

“The journalist of the future must not merely be a reporter. He or she must also be a verifier,” he said, urging journalists to establish the origins of photographs, authenticate audio recordings, examine videos and seek independent evidence before publication.

He maintained that artificial intelligence itself was not the enemy, noting that the technology could assist journalists in analysing information, help fact-checkers identify patterns, support research and improve communication.

“The question is therefore not, ‘Should we stop AI?’ We cannot. The better question is, ‘How do we govern, deploy and use AI responsibly?’” he said.

In his welcome address, the Vice-Chancellor of Caleb University, Prof. Olalekan Asikhia, said Nigeria’s expanding digital population had increased the urgency of addressing misinformation.

He cited Nigerian Communications Commission figures showing that active internet subscriptions rose from 140.3 million in August 2025 to 157.7 million in July 2026, an increase of more than 17 million in less than a year.

Asikhia also cited the Reuters Institute Digital News Report 2025, which found that 84 per cent of Nigerian respondents surveyed were concerned about distinguishing real information from fake content online, the highest proportion among the 48 countries covered.

He said the university selected the lecture theme because the problem had become central to public trust and democratic life.

“What used to be an academic habit is now a daily need for every citizen with a phone,” he said, referring to the university’s responsibility to teach students to interrogate information and establish evidence.

The vice-chancellor also identified the National Artificial Intelligence Strategy for 2025 to 2029 as part of the Federal Government’s response, but said policy commitments required research, teaching and public education to become effective.

Stay informed and ahead of the curve! Follow The Punch Newspaper on WhatsApp for real-time updates, breaking news, and exclusive content. Don’t miss a headline – join now!

Stay in the know—fast. Get instant alerts, breaking headlines, and exclusive stories with the Punch News App. Download now and never miss a beat.