EU AI law gains committee support ahead of Parliament vote


The European Parliament’s Committees on Civil Liberties (LIBE) and the Internal Market (IMCO) have overwhelmingly approved a bill establishing a risk-based framework to regulate applications of artificial intelligence. In a vote this morning, committees voted by 71 votes to 8 (with 7 abstentions) in favor of the compromise negotiated with EU member states during the late last year, as part of lengthy three-way “trilogue” talks.

The European AI law, initially proposed by the Commission back in April 2021, sets rules for AI developers based on the power of their models and/or the purpose for which they plan to apply AI. It includes a list of prohibited uses of AI (such as social scoring), as well as rules for a defined set of high-risk uses (for example, education, health or employment), such as obligations in areas such as data quality, testing and risk assessment. Additionally, it applies certain transparency requirements to general-purpose AI and tools such as deepfakes and AI chatbots.

Beyond that, most AI applications will be considered “low risk,” meaning they fall outside the scope of the law. The plan also calls for the establishment of regulatory sandboxes at the national level to allow developers to develop, train and test risky applications in a supervised “real world” environment.

The Commission’s proposal for an AI regulation did not make much noise when the EU presented it three years ago. But with the rise of generative AI over the past year, the plan has attracted worldwide attention — and driven wide divisions among the bloc’s lawmakers. MEPs proposed amending the proposal to ensure it applies to powerful general-purpose AI.while a handful of member states, led by France, have pushed in the opposition direction – seeking regulatory exclusion for advanced AI. hoping to promote national champions.

Marathon trilogue discussions in December delivered a compromise text that still included some provisions regarding general-purpose AI, leading to persistent opposition from some governments. And as recently as last month, these divisions seemed likely to derail passage of the bill. But after a critical vote by member states on the adopted compromise text earlier this month the bloc appears almost certain to adopt its flagship AI regulation within months.

That said, the bill still has a few hurdles to clear before it is adopted: there will be a plenary vote in Parliament in the coming weeks, during which MPs will be asked to formally adopt it. Then there will be a final approval from the Council.

However, these latter steps seem less likely to lead to disagreements between European co-legislators. Such a move would be a wrecking ball for the bill in the current cycle, with legislative elections looming and the end of the current college’s term – meaning legislative time and reputational room to maneuver are tight .

The full support given today by the two parliamentary committees, which have participated in the detailed examination of the legislative proposal over the years, also gives a strong signal that MPs will follow through on the required absolute majority support, which would pave the way for the adoption of the law. and will come into force later this year. The first provisions (prohibitions on prohibited practices) would then apply six months later (so probably in the second half of this year).

The EU has opted for a gradual rollout of the law, which will likely result in increased legal requirements for affected developers between 2024 and 2027. (The EU also allows nine months after a code of practice comes into force for it to apply; 12 months after entry into force for general purpose AI rules, including governance, to apply. It also states that the law will be fully applicable 24 months after its entry into force – although the obligations for high-risk systems still have a longer period of application (of 36 months).

Even if the vote in the plenary session of Parliament seems likely to be adopted, some opposition persists. The Pirate Party, for example, refuses to support what its MEPs, who today represent a few votes in committee, are doing. against the law – characterize a law as “defective”. In a statement, Pirate Party MEP Marcel Kolaja, also a member of the IMCO committee, said: “Unfortunately, despite the good position of the European Parliament, national governments have managed to paralyze the AI ​​law. Therefore, the Pirates cannot support him.

In a separate statement, Pirate Party MEP and LIBE committee member Patrick Breyer also warned: The European AI law opens the door to permanent facial surveillance in real time: more than 6,000 people are wanted by a European arrest warrant for the offenses listed in the AI ​​law. Any public space in Europe can be placed under permanent mass biometric surveillance on these bases. This law legitimizes and normalizes a culture of distrust. This is leading Europe towards a dystopian future, that of a high-tech and distrustful surveillance state.”


Leave a Comment

Your email address will not be published. Required fields are marked *