European lawmakers face a race against time to reach an agreement on new artificial intelligence (AI) rules. According to sources familiar with the matter, several issues still need to be resolved before a deal can be made, pushing any potential agreement off the table until December. The draft AI rules must be agreed upon by both the European Parliament and the member states of the European Union.
The discussions have already seen three rounds of trilogue meetings, which involve negotiations between the parliament and the EU states to finalize the laws. Another trilogue meeting is scheduled for Tuesday, following EU lawmakers' discussions on their negotiating stance concerning foundation models and high-risk AI systems. Foundation models refer to AI systems like OpenAI’s ChatGPT, which are trained on extensive datasets and have the ability to learn and perform a range of tasks.
Spain, which currently holds the EU presidency until December, has been working towards a deal and has proposed compromises to expedite the process. One of their suggestions is a tiered approach to regulate foundation models based on the number of users, with a threshold of more than 45 million users. Spain also wants additional obligations for very capable foundation models, such as ChatGPT, including regular vetting to uncover any potential vulnerabilities. However, opponents argue that smaller platforms can also pose significant risks.
Despite Spain’s efforts to find common ground, a final agreement is unlikely to be reached in the upcoming meeting. A fifth trilogue is set for early December, but if a deal is not reached then, negotiations may be further delayed by the European parliament elections in June.
Key figures involved in the process have expressed their hopes for the draft AI Act to be approved before the end of the year. EU industry chief Thierry Breton and the co-rapporteurs for the AI Act, Dragoș Tudorache and Brando Benifei, are among those who have voiced their desires for a timely approval. It is worth noting that the EU began working on the draft AI Act in 2021, and in May of this year, the European parliament agreed on draft legislation that included new regulations for facial recognition, biometric surveillance, and other AI applications. The proposed rules classify AI tools based on their perceived level of risk, with corresponding obligations for governments and companies using them.
The discussions around AI regulations in the EU highlight the complexity and importance of finding a balance between innovation and ensuring public safety. As AI technology continues to advance rapidly, policymakers and legislators face the challenging task of creating regulations that foster innovation while protecting individuals and society as a whole. Only time will tell if EU lawmakers can navigate this delicate balancing act and successfully reach an agreement on the AI rules.
Use the share button below if you liked it.