EU’s AI regulation faces uncertainty as deal eludes leaders

According to AP reports, the European Union’s ambitious plans to lead the world in AI regulation are facing uncertainty as leaders struggle to reach a deal. The EU’s Artificial Intelligence Act, considered a groundbreaking rulebook, may face disruption due to the rapid development of generative AI systems like OpenAI’s ChatGPT.

Governments worldwide have been rushing to regulate AI technology, with concerns about its risks and potential impact. The United States, the United Kingdom, China, and global coalitions like the Group of 7 have all joined the race to catch up with Europe in regulating this emerging field.

Apart from addressing generative AI, EU negotiators also need to tackle other contentious issues such as limitations on AI-powered facial recognition and surveillance systems, which have raised privacy concerns.

While there is optimism about reaching a political agreement, there are significant obstacles that could hinder the deal. Kris Shrishak, a senior fellow specialising in AI governance at the Irish Council for Civil Liberties, stated that although negotiators are eager for a political win, the critical nature of the issues at hand means a deal is not guaranteed.

Carme Artigas, Spain’s AI and digitalisation minister, mentioned that 85 percent of the technical wording in the bill has already been agreed upon. However, if a deal is not reached in the current round of talks, negotiations may be delayed until next year. This raises the possibility that the legislation could face further delays or take a different direction under new leadership after the EU-wide elections in June.

One of the major points of contention is the regulation of foundation models, which are the advanced systems that underpin general-purpose AI services. These models, such as OpenAI’s ChatGPT and Google’s Bard chatbot, have raised concerns among researchers who fear their potential misuse for disinformation, cyberattacks, or bioweapon creation.

France, Germany, and Italy have resisted the inclusion of foundation models in the legislation, advocating instead for self-regulation. This change of heart is seen as an attempt to support domestic generative AI players and compete with major U.S. tech companies like OpenAI.

Brando Benifei, an Italian member of the European Parliament leading the negotiating efforts, expressed optimism about finding common ground with member states. While progress has been made regarding foundation models, challenges remain in reaching an agreement on facial recognition systems.