The EU's fight against deepfakes and disinformation ahead of elections

Policymakers in democracies across Europe and around the world have made countless political promises to fight disinformation, but this battle cannot be won if we continue to operate in national silos.

Disinformation and deep fakes – content manipulated to appear real – can pose a threat to our democracy if used by bad actors to deceive the public. These fabricated pictures, audio clips, or videos can erode trust in institutions, sow discord in public discourse, and have a destabilising role in elections.

Policymakers and tech companies are on the same side in this fight. We agree that more needs to be done. But we have the right rules already in place; now we need to learn how to deploy them and make them work. ?

Companies are on the front lines of this battle

The tech sector is investing in top-notch technology and automated detection tools to identify and label AI-generated content and ultimately remove harmful content. For example, companies use machine learning algorithms to analyse posts, comments, and images and flag content that violates community standards for review by human moderators. Fact-checking initiatives are also booming with partnerships between social media platforms and reputable news organisations. Moreover, media literacy programmes are gaining traction, aiming to equip users with the critical thinking skills necessary to differentiate facts from fiction online.

We have to pull our weight to remain one step ahead of bad actors

There is no silver bullet, and bad actors are constantly learning and evolving. Disinformation campaigns are becoming cheaper to run and more sophisticated, exploiting tech advancements such as AI. Bad actors learn from platforms’ detection strategies and are getting better at sidestepping them.

Rules such as the Digital Services Act, Political Ads Regulation, and the AI Act and tools such as the DSA Election Guidelines, the Code of Practice on Disinformation, the AI Pact, and the European Democracy Action Plan are all promising initiatives that address disinformation. Yet their requirements can be unclear and, at times, even contradictory, and governments and companies are learning how to deploy them.

We need to make sure all stakeholders, including authorities and businesses, are equipped with the skills and knowledge to fight disinformation and use the existing tools available. The AI Act, for example, about to enter into force, will require AI providers and deployers to clearly label deepfakes. If the deep fakes are not properly marked, Member States authorities can issue fines, which can go up to 3% of the company's total worldwide annual turnover.

President Ursula von der Leyen's proposed the European Democracy Shield which has the potential to offer a path forward. This initiative could be very impactful if it manages to draw clear responsibilities and ensure coordination between Member States and EU authorities rather than adding new rules and governance layers. The new structure should have a single point of reference for tech companies. The Democracy Shield could also bring together public and private sector tech talent to fight disinformation and develop a major reskilling programme for the legal system and law enforcement, so they have the knowledge and understanding to detect and tackle these crimes.

Finally, it's important for the EU to clearly lay out its priorities on this issue, making sure that the different initiatives and laws work well together. Platforms do have a role in fighting disinformation and deep fakes, but a clear and unified framework from the EU is crucial. By collaborating, we can protect online spaces and ensure a healthy, democratic digital world.

Disinformation is a tricky problem, but it's not unbeatable. By working together, we can create a more informed and resilient digital Europe.