TL;DR: The UK Home Office is working with Microsoft to develop a deepfake detection evaluation framework. The initiative brings together technology companies and academia to establish industry standards for identifying AI-generated fakes, with deepfake volumes rising from 500,000 in 2023 to 8 million in 2025.

Scale of the Problem

The UK government announced plans on Thursday to implement a structured approach to evaluating deepfake detection tools, with Microsoft as a key partner. The framework will draw on expertise from technology firms and academic researchers to assess how effectively current tools can identify and decode deepfake content.

The numbers behind the initiative are stark. According to the Home Office, approximately 8 million deepfakes were shared in 2025, up from 500,000 just two years earlier. Deepfakes have been used in crypto scams impersonating public figures including Sir Keir Starmer and Prince William, as well as in cases of sexual abuse, identity fraud, and corporate impersonation.

Tech Secretary Liz Kendall described the threat plainly: “Deepfakes are being weaponised by criminals to defraud the public, exploit women and girls, and undermine trust in what we see and hear.”

The government has already made non-consensual sexually explicit deepfakes a criminal offence. Safeguarding Minister Jess Phillips said the framework would help “close loopholes” and hold the technology industry to account.

Andrea Simon, director of the End Violence Against Women Coalition, welcomed the move but cautioned that the burden should not fall on victims. “The onus cannot be on victims to detect and report abuse and battle with platforms to have this violating material taken down,” she said. “It is the platforms themselves who could and should be doing so much more.”

Looking Forward

The framework aims to set detection standards that the wider industry can adopt. With deepfake volumes growing at pace and the technology becoming more accessible, establishing reliable detection benchmarks is becoming a practical necessity for businesses and public institutions alike. Singapore issued similar warnings to businesses about corporate deepfake scams in 2025, suggesting this is a challenge no single country can address in isolation.

Share this article