Big tech told to identify and label AI deepfakes ahead of EU elections

The European Union has urged Facebook, TikTok, and other major tech companies to take decisive action against deepfakes and other AI-generated content by implementing clear labels ahead of the European elections in June.

The move is part of a string of measures published under the EU's landmark AI legislation, approved by lawmakers earlier this month, that governs AI content on 22 "very large” social sites including Snapchat, Instagram, YouTube and X.

The European Commission wants those sites to clearly label political advertising and reduce "virality of content" and disinformation that could jeopardise the integrity of the polls.

Brussels is particularly wary about the impact of Russian "manipulation".

Risk assessment

In the newly released guidelines, the EU Commission said major platforms must "assess and mitigate specific risks associated with AI, such as by clearly labelling content generated by AI, otherwise known as deepfakes.

"With today's guidelines we are making full use of all the tools offered by the DSA <[a href="https://digital-strategy.ec.europa.eu/en/policies/digital-services-act-package#:~:text=The Digital Services Act and,level playing field for businesses." rel="nofollow,noreferrer,noopener" target="_blank">Digital Services Act] to ensure platforms comply with their obligations and are not misused to manipulate our elections, while safeguarding freedom of expression," said Europe's digital Commissioner Thierry Breton.

While the guidelines are not legally binding, platforms are required to outline alternative "equally effective" measures if they choose not to comply.

The EU may request further information if regulators do not believe there is non-compliance, while companies also risk being investigated and fined.

The EU has said it will conduct "stress-tests" in collaboration with relevant platforms in late April to assess their preparedness.

X has already been under investigation since December over content moderation.

Earlier this month the commission asked Facebook, Instagram, TikTok and four other platforms to provide more information on how they are countering AI risks to polls.

(with newswires)