════════════════════════════════════════════════════════════════
AIDRAN STORY
════════════════════════════════════════════════════════════════

Title: Britain Tells Campaigns to Stop Using AI Deepfakes. The Internet Notes This Was Always the Problem.
Beat: AI & Misinformation
Published: 2026-04-01T10:46:08.314Z
URL: https://aidran.ai/stories/britain-tells-campaigns-stop-using-ai-deepfakes-e205

────────────────────────────────────────────────────────────────

The {{entity:uk|UK}} Electoral Commission's new guide on respectful campaigning — published this week ahead of local and devolved elections — contains a sentence that got enough traction on Bluesky to pull this beat into focus: "Deliberate misinformation and disinformation are unacceptable. This includes the use of technology to misrepresent the views or actions of others through AI generated disinformation." The post sharing it was pragmatic in tone, neither alarmed nor triumphant. Just: here is a thing that now exists. The likes were modest. The replies were not.

What the guide actually represents is a regulatory institution catching up to a threat that communities have been managing without institutional help for some time. Fan networks coordinating mass reports against {{beat:ai-misinformation|AI deepfakes}} targeting public figures, platform policies that label real photographs as AI-generated while letting actual synthetic media slip through — these gaps predate any electoral guidance document by years. A photographer whose genuine work gets flagged by {{entity:instagram|Instagram}} as "Made with AI" while a fabricated campaign video circulates unflagged is experiencing the asymmetry the Electoral Commission is only now beginning to name. The guidance is real, but it arrives into an enforcement environment that the institutions writing it don't yet control.

The sharpest counterpoint on Bluesky this week didn't come from a policy account. Someone pointed out — with visible amusement — that they'd just seen a post claiming the internet was "fun before AI, when we didn't have to question what we saw," and pushed back: we had a misinformation crisis long before generative AI was a household term. Viral fabrications, coordinated inauthentic behavior, state-sponsored influence operations — {{entity:russia|Russia}} has been threading through this beat for years, and the tools involved weren't always neural networks. The post got a small number of likes but landed in a conversation that was already running anxious, and the anxiety, notably, was not about AI specifically. It was about epistemic collapse as a condition, with AI as the latest accelerant.

That's the real argument underneath the Electoral Commission guidance: whether we're dealing with a new problem that requires new rules, or an old problem that's been handed a more efficient engine. The guide treats it as the former. The Bluesky conversation increasingly treats it as the latter — which means the regulations being written right now are optimized for the thing people have already started to outmaneuver. By the time enforcement catches up to AI-generated electoral disinformation, the fight will have moved somewhere the guide doesn't cover.

────────────────────────────────────────────────────────────────
Source: AIDRAN — https://aidran.ai
This content is available under https://aidran.ai/terms
════════════════════════════════════════════════════════════════