The Artificial Intelligence (AI) Act, which would be the first set of comprehensive laws relating to AI regulation, has been passed by EU legislators following months of discussion and two years after draught regulations were originally put up. The trilogue is the following phase, during which member states and EU legislators are going to negotiate over the final specifics of the law.
Members of the European Parliament (MEPs), a subset of “General Purpose AI” that encompasses products like ChatGPT, reportedly agreed to prior recommendations to impose stronger regulations on foundation models. Companies who provide ChatGPT and other generative AI tools would be required, under the recommendations, to disclose whether they have integrated any copyrighted content into their systems.
According to the source, the AI Act’s draught saw one notable last-minute alteration that had to do with generative AI models. These models would need to be created and developed in compliance with EU legislation and fundamental rights, such as freedom of expression.
According to Mher Hakobyan, Amnesty International’s campaign adviser on AI legislation, the AI Act presents EU legislators with a chance to end the deployment of discriminatory and human rights-infringing artificial intelligence (AI) systems.
Many people have been waiting for the government’s AI regulation
Many state-based AI-related measures have been approved in the U.S., but the EU AI Act is the major government regulation that many in the AI and legal communities have been waiting for.
Avi Gesser, a partner at Debevoise and Plimpton and cochair of the firm’s cybersecurity, privacy, and artificial intelligence practice group, stated in December that the AI Act aims to create a risk-based regime to address the outcomes of AI that pose the greatest risk while striking a balance so that laws do not stifle innovation.
It’s about realizing that some use cases will be low-risk and won’t necessitate a lot of regulation, he said. He clarified that the EU AI Act will serve as an example of a comprehensive European law going into effect and gradually permeating numerous state- and sector-specific regulations in the United States, similar to the privacy-focused GDPR.
The AI Act will have a global impact because it will apply to businesses that provide or use AI systems in the EU as well as users or providers of AI systems in third countries (including the UK and the US), provided that the output of those systems is used in the EU.
