Jun 3, 2026

The U.K. has mandated that Google provide publishers the option to opt out of AI search integration. This regulation gives publishers more control over how their content is used, potentially strengthening their bargaining power in content deals with Google. Google plans to test this feature with select U.K. publishers before expanding it globally.

The Canadian federal government plans to unveil its new artificial intelligence strategy soon. Simon Fraser University experts are available to discuss the strategy's implications, including ethics, youth protection, and data sovereignty. Their insights may help inform public understanding of the government's approach to AI challenges.

Lovable signed a multiyear deal with Google Cloud to increase its usage fivefold, including AI resources. This partnership will help Lovable expand its footprint among enterprise customers while supporting Google’s investment in the AI model Anthropic. The agreement also allows Lovable’s agents to be sold through Google Cloud’s marketplace, simplifying enterprise procurement.

The Trump administration issued an executive order on June 3, 2026, that requires AI models to undergo national security vetting 30 days before public release. This move attempts to balance the need for AI innovation with growing national security risks posed by advanced technologies. Experts question the effectiveness of this voluntary arrangement, noting it may leave security decisions in the hands of a few private companies.

A study by Penn State researchers found that AI chatbots, like ChatGPT, respond to healthcare queries with nearly 76% accuracy. This level of accuracy raises concerns about the reliability of AI in real-world medical situations, suggesting that trained physicians are better suited to handle patient inquiries. The team will present their findings at the Association for Computing Machinery Fairness, Accountability and Transparency conference later this month.

Researchers Travis Weber and Rohit Taneja introduced the Digital Apprentice framework for developing agentic AI. This framework aims to balance autonomy and oversight, allowing AI systems to learn and earn independence under human guidance. It emphasizes safety and accountability, potentially shaping future practices in AI deployment.

Researchers introduced a new framework for verifying enterprise artificial intelligence agents before deployment. This approach improves regulatory coverage and scenario generation, addressing gaps in post-deployment monitoring. The study's results suggest a viable way to enhance trust in AI systems across regulated industries.

Researchers have developed a benchmark called Curation-Bench to test if generalist coding agents can automate data curation for AI. The study finds that while these agents can run the curation loop effectively, they require structured guidance to explore new data policies beyond basic enhancements. This suggests that future advancements in automated data curation will depend on improved methods for guiding agent exploration.