Aaron Mak

Days after Sam Altman and other AI executives joined calls for a slowdown of artificial intelligence development, OpenAI disclosed additional instances of its models going rogue over the past six months. The company published a blog post last week detailing six instances of models acting in “unexpected or concerning” ways since April.

Researchers at the Stanford Institute for Human-Centered AI and the university’s RegLab recently published a study demonstrating how large language models can help identify discriminatory local laws. The AI system helped human reviewers identify thousands of local statutes that could violate anti-discrimination laws by treating people differently based on race, gender, and citizenship. HAI fellow Daniel E. Ho, who helped conduct the study, said AI is well-suited for the task because of the sheer scale of local laws.

The increasingly urgent pleas from leading tech executives to slow the advance of artificial intelligence are confronting one daunting obstacle: It’s not entirely clear how they could do that. A slowdown could involve releasing frontier models less frequently, requiring approval before deploying systems with powerful cyber or biological capabilities, giving independent monitors access to AI labs, or negotiating limits with China.

President Donald Trump is once again expanding the frontiers of social media law. Truth Social began offering a $100,000-per-month subscription for early access to Trump’s posts in July. The Trump Media and Technology Group, Truth Social’s parent company, advertised the service as a “real-time feed of the platform’s most market-moving” content.

The technolibertarian dream of a borderless internet may soon get its biggest reality check in years.

There’s been a lot of concern over what AI chatbots communicate to users—from conversations that may encourage unhealthy emotional attachments, to advice on committing crimes. Much of the debate over regulatory proposals and lawsuits concerning chatbots has focused on speech rights. There are open questions over whether outputs can be considered a form of legally protected speech. Yet an overlooked aspect of these issues is the potential right of users to access chatbot outputs. The First Amendment protects not only the right to speak, but also the right to listen.

While social media bans for kids have gained a foothold in more than 40 countries, it’s still an open question whether they’re legal in the US.

On April 24, the Justice Department intervened in xAI’s challenge to Colorado’s “Consumer Protections for Artificial Intelligence” law. In its complaint, the DOJ argues the law’s provisions curbing algorithmic bias violates people’s 14th Amendment right to be treated equally under the law.

Long a hobbyhorse for futurists and sci-fi authors, the prospect of a catastrophe precipitated by AI has lawmakers asking who would pay for the damage. Wired recently reported that OpenAI is backing an Illinois state bill that would protect frontier AI companies from liability for causing serious widespread injury that affects more than 100 people or does more than $1 billion in property damage, such as in the case of a chemical, biological or nuclear catastrophe.

The 2017 “alternative facts” statement made by the White House was a turning point. We are now in a place where the person who gets their version of the truth out fastest and biggest wins. The fact that President Trump has a social network he calls “Truth Social” is not an accident. He wants to own the word “truth.” We’re past the point where there’s a singular truth or a singular fact.