Our latest threat report examines how malicious actors combine AI models with websites and social platforms—and what it means for detecti...
OpenAI appoints Arvind KC as Chief People Officer to help scale the company, strengthen its culture, and lead how work evolves in the age...
SWE-bench Verified is increasingly contaminated and mismeasures frontier coding progress. Our analysis shows flawed tests and training le...
OpenAI announces Frontier Alliance Partners to help enterprises move from AI pilots to production with secure, scalable agent deployments.
Voices contain countless cues about their owners, and new research suggests that computers might use them t...
AI could soon spew out hundreds of mathematical proofs that look "right" but contain hidden flaws, or proof...
We share our AI model’s proof attempts for the First Proof math challenge, testing research-grade reasoning on expert-level problems.
Qubits, the heart of quantum computers, can change performance in fractions of a second — but until now, scientists couldn’t see it happe...
Research from the MIT Center for Constructive Communication finds leading AI models perform worse for users with lower English proficienc...
The framework predicts how proteins will function with several interacting mutations and finds combinations that work well together.
A new method developed at MIT could root out vulnerabilities and improve LLM safety and performance.
3.1 Pro is designed for tasks where a simple answer isn’t enough.
OpenAI commits $7.5M to The Alignment Project to fund independent AI alignment research, strengthening global efforts to address AGI safe...
By minimizing the need to drive around looking for a parking spot, this technique can save drivers up to 35 minutes — and give them a rea...
OpenAI for India expands AI access across the country—building local infrastructure, powering enterprises, and advancing workforce skills.
The Gemini app now features our most advanced music generation model Lyria 3, empowering anyone to make 30-second tracks using text or im...
Some say we’ve entered a new age of AI-enabled scientific discovery. But human insight and creativity still can’t be automated.
The context of long-term conversations can cause an LLM to begin mirroring the user’s viewpoints, possibly reducing accuracy or creating ...
OpenAI and Paradigm introduce EVMbench, a benchmark evaluating AI agents’ ability to detect, patch, and exploit high-severity smart contr...
Machine Perception
Subtle shifts in how users described symptoms to AI chatbots led to dramatically different, sometimes dangerous medical advice.
Google DeepMind brings National Partnerships for AI initiative to India, scaling AI for science and education
Neuromorphic computers modeled after the human brain can now solve the complex equations behind physics simulations — something once thou...
A researcher from the University of Essex dives into the philosophical and ethical questions surrounding "d...