AI & Disinformation: The Eroding Trust in News & Reality

How the rapid advancement of generative AI is fundamentally altering our relationship with information and challenging our ability to distinguish truth from fabrication.

Published on Feb. 15, 2026

Got story updates? Submit your updates here. ›

The rapid advancement of generative AI is not just changing how we create content, but is fundamentally altering our relationship with information itself. Even established institutions are struggling to keep pace, as the sophistication of AI-generated content now surpasses our ability to reliably verify visual evidence. The mass adoption of these technologies, coupled with the proliferation of deliberate disinformation campaigns and the 'slopification' of social media, is eroding public trust in news and reality.

Why it matters

The erosion of trust in news and information has far-reaching implications for society. It empowers the spread of misinformation and extremist ideologies, weakens critical thinking skills, and poses a significant challenge to democratic institutions and processes. Addressing this crisis requires a multi-faceted approach, including enhancing media literacy, developing robust technological solutions, and countering foreign interference.

The details

The challenge isn't limited to individual consumers, as even established institutions like the New York Times have struggled to verify the authenticity of visual evidence provided by sources. A BBC study revealed that 45% of chatbot responses contained at least one significantly misleading statement, highlighting the potential for widespread misinformation. Networks aligned with Russia and those supporting Donald Trump are increasingly targeting France with differing methods but a converging narrative, posing an 'unprecedented challenge.' The proliferation of low-quality, AI-generated content, known as 'slopification,' is further exacerbating the problem by desensitizing people to falsehoods and weakening critical thinking skills.

  • In January 2026, the New York Times was forced to retract a photograph of Nicolas Maduro's arrest provided by the White House due to concerns about its authenticity.
  • A BBC study from October 2025 revealed that 45% of chatbot responses contained at least one significantly misleading statement.

The players

William Audureau

A journalist at Le Monde who aptly described the current situation as a world of 'fakes and truths.'

Thomas Renault

A lecturer at Paris-I who explains that the issue isn't simply about occasional mistakes, but the potential for widespread misinformation.

Nicolas Hénin

A researcher at the University of Manchester who believes that 'slopification' desensitizes people to falsehoods and weakens their critical thinking skills.

Sora

An AI system that created videos that ChatGPT failed to identify correctly 92.5% of the time.

Grok

An AI system that created videos that failed to be identified correctly 95% of the time by ChatGPT.

Got photos? Submit your photos here. ›

What they’re saying

“You can no longer trust your eyes.”

— William Audureau, Journalist (Le Monde)

“The issue isn't simply about occasional mistakes, but the potential for widespread misinformation.”

— Thomas Renault, Lecturer (Paris-I)

“Slopification desensitizes us to falsehoods and weakens our critical thinking skills.”

— Nicolas Hénin, Researcher (University of Manchester)

What’s next

The French Response account on X, the foreign ministry's platform for countering disinformation, represents an innovative approach, effectively challenging foreign interference. However, its long-term effectiveness remains an open question.

The takeaway

The current situation demands a multi-faceted response, including increased media literacy, technological solutions such as robust watermarking and authentication systems, and responsible development and deployment of these technologies to avoid unintended consequences like censorship or the suppression of legitimate expression.