After two days of intense negotiations, the European Parliament, member states and the Commission reached an agreement on the world's first-ever regulation on AI. This is a major victory for the Socialists and Democrats, who have worked relentlessly to ensure that the world’s first regulation on AI takes fundamental rights into account.
The S&Ds played a crucial role by ensuring the introduction of Fundamental Rights Impact Assessments (FRIA). This means that companies using high-risk AI systems will have to carry out an assessment to identify risks associated with the specific context of deployment of AI systems, the people that could be affected, and also to identify measures to mitigate such risks.
Among the key priorities the S&Ds managed to anchor in the text are: extending the scope to protect democracy, rule of law and the environment; obliging providers to ensure a sufficient level of AI literacy of their staff; obliging companies using high-risk AI systems at the workplace to inform workers’ representatives and the affected workers before such AI systems are put into service – among many others.
Brando Benifei, S&D MEP and European Parliament co-rapporteur on the Artificial Intelligence Act, said:
“Having reached an agreement is a great victory for all of us. Today, we show that is possible to stimulate innovation while taking care of people’s wellbeing.
“Setting clear rules for powerful foundation AI models is a clear sign that Europe is putting safety over the interests of the most powerful. While it is true that generative AI systems can enhance creativity, they can also create the spread of fake news or deep fakes that can harm people and endanger our democratic debates. The European Parliament has been clear and ambitious on this point: we really wanted enforceability; we really wanted to ensure that they could not escape them.
“Artificial Intelligence will definitely impact and define the workplace and the job market. It is almost impossible to stop this trend. However, the application of AI at the workplace and its effect on employment cannot be fully addressed in the AI Act. Solving this issue requires more targeted and specific initiatives such as a specific EU legislation on AI at the workplace. This is why, even if it does not specifically regulate workplace-related AI uses, the new rules recognise the need for such regulation.”
Next steps
The Parliament and the Council will formally adopt the agreed text to become EU law. The EP Internal Market and Consumer Protection (IMCO) and Civil Liberties, Justice and Home Affairs (LIBE) committees will vote on the agreement in their upcoming meetings.
