In autumn 2025, British Conservative MP George Freeman became the victim of an AI-generated deepfake video claiming he had defected to the rival Reform party. The claim was entirely fabricated — but plausible enough, as several real Conservative defections had taken place.
The Confrontation
In March 2026, Freeman had the rare opportunity to confront representatives from Meta, Google and X directly in the British Parliament. The result was sobering.
Google representative Zoe Darme explained the policies on election advertising and "violative" content but could not say whether something "demonstrably false" constitutes a violation per se.
X representative Wifredo Fernandez referenced the policy on "synthetic and manipulated media" but conceded that the three-step test might not have applied.
Meta, Google and X could neither explain why the video circulated for so long nor guarantee it would not happen again.
The Systemic Problem
Freeman argued that uncontrolled AI-generated political disinformation could end the democratic process in Britain. "There was no legal recourse. There was no explanation or principle that it was a problem," he told the Science, Innovation and Technology Committee.
The hearing reveals a pattern: platforms have policies but no functioning mechanisms to detect and remove AI-generated political disinformation in real time.
Sources: BBC, The Register (26.03.2026), UK Parliament Science, Innovation and Technology Committee.