The EU just announced that it negotiated a political deal on the Artificial Intelligence Act, leading the way on the first comprehensive AI law in the world. It reached an agreement on controversial topics such as predictive policing, facial recognition and the use of AI by law enforcement. More technical work will flush out specifics but this a huge development providing much needed guidance to companies.
During last month’s AI Transformation Leadership Council, I provided updates on global AI regulation. Although this tech is evolving rapidly, legal frameworks are struggling to keep pace. There is no single global framework, but instead, there is a patchwork of national and regional approaches. Below are highlights:
The EU's AI Act regulates the sale and use of AI systems based on a risk-based framework, with categories for unacceptable, high, limited, and minimal risk. Unacceptable risk systems, considered a clear threat to safety, livelihoods, and rights, will be banned. While high risk systems such as those used in facial recognition and recruitment, will be subject to strict requirements, including transparency, human oversight, and data governance.
In the US, Congress hasn't passed national #ai or data privacy laws even though over 50 AI and data privacy bills have been introduced during this 118th Congressional session. Despite this, Congress is not likely to pass a broad national AI law anytime soon. Any successful laws will likely focus on less controversial issues like funding AI research and domain-specific agency actions especially in health care, financial services, housing, workforce, and child safety.
As a result of Congress’ lack of progress on AI regulation, states are acting through their legislatures and governor executive orders (14 states enacted AI laws in the past year).
President Biden issued an Executive Order on safe, secure, and trustworthy AI in October with guidelines for development and use. He used the Defense Production Act as the basis for the Order. Now, U.S. government departments/agencies will work on implementation, which focuses on:
- Ensuring AI safety and security through testing standards and public-private collaboration.
- Protecting privacy via supporting privacy-enhancing technologies.
- Advancing equity by tackling algorithmic discrimination in areas like criminal justice.
- Empowering consumers and workers through policies curbing AI harms.
- Driving competition and innovation via research investments and immigration reforms.
It is important to note that unlike acts of Congress, Executive Orders cannot create new agencies or grant new regulatory powers over private companies. They can also be undone by the next president.
Overall, a uniform global framework is needed to ensure AI safety and responsible development. A patchwork of regulations creates confusion and challenges for businesses. #airegulation #artificialintelligence #tech
https://lnkd.in/ezWT7ZPa
Data Privacy and Cybersecurity Counsel
3moThanks for flagging!