MEPs to draft additional AI oversight measures

A group of 12 Members of the European Parliament will draft a set of rules to oversee the development of AI systems. This is reported by Reuters.

The lawmakers working on a comprehensive AI act said that technological progress “faster and more unpredictable” than expected.

“The recent emergence and broad public access to powerful AI, together with exponential improvements in the performance of generative algorithms, have prompted us to pause and reflect on our work,” the open letter from European Parliament members states.

They say that an additional set of preparatory rules is also necessary to regulate the growth and deployment of powerful general‑purpose AI.

The signatories urged the head of the European Commission Ursula von der Leyen and U.S. President Joe Biden to hold a global AI summit. They believe the event would help world leaders agree on a set of guiding principles aimed at the development and deployment of powerful AI.

MEPs also urged companies and laboratories to demonstrate “a high sense of responsibility”, increase transparency and cooperate with policymakers.

The lawmakers warned that political inaction “could widen the gap between AI development and the ability to govern it”. They called for mobilising industry, research and decision-makers in Europe and worldwide.

Meanwhile, the EU’s AI Act has been in draft form for nearly two years. According to media reports, the document is currently being discussed by a parliamentary committee, which hopes to reach a common position by 26 April.

Some MEPs consider some statements in the letter to be overly panicky. However, they agree with its core idea: the rapid evolution of powerful AI requires political attention to prevent complex future scenarios.

As noted, in late March Italian authorities blocked the operation of ChatGPT on the territory of the country due to data privacy concerns. Other European countries have not ruled out adopting similar measures.

In the same month, a group of over 1,000 experts urged to pause the development of language models more powerful than GPT-4.