This week’s AI policy update.

First, new regulation: California AI law

This week, California Governor Gavin Newsom signed the Transparency in Frontier AI Act into law.

The law is the first of its kind in the US and regulates frontier AI firms, requiring them to fulfil transparency commitments and report AI-related safety incidents.

NBC AI reporter Jared Perlo has more here: https://lnkd.in/eC4zxcvw.

Meanwhile, Jonas Freund asks how the new law attempts to address key risks and how whistleblower protections change accountability. Read his insightful post here: https://lnkd.in/e3zDKVNX

Next, new report: Is the UK Government prepared for a major AI incident?

The Centre for Long-Term Resilience has published a new report this week, assessing whether the UK Government is prepared for major AI incidents.

The short answer is no, according to the report.

However, as the report’s author Tommy Shaffer Shane says, there are steps the Government can take to correct this failing.

See Tommy’s post here for the his recommendations: https://lnkd.in/eFAkBU5U.

Last, new thinking: why AI needs social science

In a fascinating and persuasive piece, Cosmina Dorobantu, Professor of Practice at LSE, and Helen Margetts, Professor of Society and the Internet at the Oxford Internet Institute, University of Oxford, argue why AI needs social science.

The most sophisticated AI models are about people. And if people are central to AI, they argue that the social sciences must play a crucial role in its development.

The social sciences offer a lens through which we can study and understand people. They shed light on how economies function, how politics, governance, and law shape our environment, and what drives societies and human behaviour.

Therefore, the only way to shape an AI future that serves humanity is by building bridges between AI and the social sciences, they conclude.

The piece is well worth your time.

See Cosmina’s post here: https://lnkd.in/eDhe7UJe.