Election watchdogs turn AI detectives, but UK body says bias already tainted a poll

The UK Electoral Commission is developing an AI-powered system to notify candidates of significant deepfakes within three hours of their posting to social media, chief executive Vijay Rangarajan said at a panel during the Conservative Party's annual conference in the week before 15 October 2026.

Rangarajan also warned of persistent hacking and phishing attacks against the regulator. He said the commission had detected signs of "algorithmic bias" — a form of electoral interference that would likely require the involvement of a social media company itself — in at least one UK election.

Separately, Nigeria's Independent National Electoral Commission announced on 14 October 2026 that it had established an AI unit to monitor digital threats, detect deepfakes and prevent manipulation of the 2027 general elections. INEC chairman Prof. Joash Amupitan said: "Algorithms will not, and must never, decide the 2027 vote." He warned that AI-generated videos, cloned audio recordings and other synthetic media could spread false information about candidates, election postponements and electoral procedures.

INEC pledged credible 2027 elections amid warnings from Yiaga Africa, the ECOWAS Network of Electoral Commissions and development partners that AI could manipulate voters, spread false results and undermine confidence in the polls.

The two announcements point to a shared problem. Electoral commissions are becoming de facto technology regulators, building detection tools for synthetic media while acknowledging that platform algorithms themselves may already be interfering with elections. There is no evidence the UK and Nigerian systems are coordinated.

The three-hour notification window suggests the UK regulator sees deepfakes as a speed problem: false claims can spread before corrections land. But notifying candidates is not the same as removing content or issuing a public correction. The evidence does not say who receives alerts, what they can do with them, or how much protection voters actually get.

The UK system is still in development. No launch date or completion timeline has been given, and the evidence does not say whether it was used in the Holborn and St Pancras by-election held on 8 October 2026.

Several questions remain unanswered. What specific evidence of algorithmic bias did the Electoral Commission detect, and in which UK election? Will the UK system be operational before the next general election, and has it been tested? What powers, if any, do electoral commissions have to compel social media platforms to act on flagged deepfakes?

What is clear is that two electoral bodies on different continents are racing to build AI safeguards before major votes — while one of them reports being hacked and detecting algorithmic interference in its own elections.