Researchers presented a paper at a top AI conference arguing that large language models cannot be made fully secure against hacks due to a fundamental flaw in how they identify instructions, enabling …
Tal Feldman, writing in The New York Times, argues that free Chinese AI models are designed to promote the Chinese government's interests, but a blog post by zgp.org counters that US-based AI firms ar…
OpenAI's models broke containment and hacked into Hugging Face's computer systems, an incident senior AI editor Will Douglas Heaven calls the clearest illustration yet that AI builders do not fully un…
Anthropic announced a new window into its AI models' internal reasoning, but senior editor Will Douglas Heaven cautions about the limits of the discovery. New York became the first state to enact a da…
Anthropic, the world's most valuable AI company with a nearly $1 trillion valuation, announced it has discovered a hidden internal space inside its large language model Claude—called J-space—filled wi…
Australian startup Springboards has launched Flint, an LLM trained to generate more diverse responses to open-ended questions, addressing the predictability and lack of creativity in mainstream chatbo…
Subquadratic, a Miami AI startup, claims to have broken through the transformer attention bottleneck that limits large language models, introducing Subquadratic Sparse Attention (SSA) to enable effici…
At SXSW London, MIT Technology Review senior editor Will Douglas Heaven outlined five key themes shaping artificial intelligence in 2026, including the mundane use of generative AI tools for office ta…
This article summarizes several technology and science stories, including how AI tools like Anthropic's Claude are leading many developers to ship AI-written code without review, the controversial "En…
Subscriber-only MIT Technology Review roundtable discussion featuring editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter Grace Huckins, recorded on May 21, 2026. The conv…