May 10, 2026

Greece is updating its constitution to ensure that artificial intelligence serves human society. Prime Minister Kyriakos Mitsotakis described these changes as essential for protecting future generations.

Google, Microsoft, and xAI will allow the US government to test their artificial intelligence models before public release. The evaluations, conducted by the Center for AI Standards and Innovation, will focus on risks related to cybersecurity, biosecurity, and chemical weapons.

The Trump administration is considering policy changes that would limit how private technology contractors can influence the government’s use of artificial intelligence. Draft documents suggest the government wants clear control over technology applications, emphasizing that elected officials should determine lawful uses rather than companies.

The Pentagon awarded a $500 million contract to Scale AI, which is backed by Meta Platforms Inc., to analyze data and support decision-making. This new deal marks a five-fold increase from a previous $100 million contract and highlights the military's growing commitment to artificial intelligence technology.

Anthropic announced a joint venture for enterprise AI services with partners including Blackstone and Goldman Sachs, valued at $1.5 billion. Hours earlier, OpenAI revealed a similar venture called The Development Company, aiming to raise $4 billion, indicating strong competition between the two firms in the enterprise AI space.

OpenAI launched GPT-5.5-Cyber, a new model designed for cybersecurity teams, shortly after Anthropic's Mythos preview. This upgrade focuses on tasks like vulnerability identification and malware analysis and is available to a wider range of vetted users than its competitor.

Models Table Pro, Anthropic 80-fold growth, OpenAI phone rumors, and much more!

Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI systems are about to start building themselves. What does that mean? I’m writing this post because when I look at all the publicly available information I […]

A new AI system called Synthegy allows chemists to design complex molecules by using simple language to guide synthesis and reaction planning. This tool improves the retrosynthesis process by combining algorithms with large language models, enabling faster and more intuitive decision-making.

We reviewed the “Risks from automated R&D” section of Anthropic’s February 2026 Risk Report , producing two corresponding review documents: our original review and our updated review . We recommend that readers refer to our original review, which represents our review of the report as originally received. We expect the public version of the report to be updated to resemble the changes anticipated in our updated review in content, though not necessarily in exact wording. We expect our updated review to cover those changes, but if the updated public version includes any changes that materially affect our opinion...

In this episode, Greg Burnham and Tom Adamczewski join Anson Ho to push back on benchmark pessimism and dig into what the next generation of AI benchmarks could look like.

GPT-5.5 Pro achieves a new high score of 159 on the Epoch Capabilities Index — our statistical tool that combines multiple benchmarks into a unified scale.

Give up at least one of: text only, short time horizon, easy to grade, and expert human superiority.

Summary: We describe three different definitions of the productivity impact of AI (AKA uplift), and show there’s reason to expect: \[\text{uplift on old tasks} \leq \text{uplift in value} \leq \text{uplift on new tasks}\] Three Measures of Uplift One complication in measuring AI’s effect on productivity is that it has different effects on different tasks, and this causes people to change how they allocate their time between tasks. This makes it more difficult to talk about the effect of AI on overall productivity. We use “old tasks” to mean the set of tasks you’d do in a typical day before AI is available – yo...

AI chip supply chain bottlenecks, smuggling to China, benchmark saturation, revenue efficiency at AI companies, and more

We estimate that between 290,000 and 1.6 million H100-equivalents (H100e) were smuggled to China through 2025. Our median estimate of 660,000 H100e would be roughly a third of China's total compute.