Research & Benchmarks - Latest AI News & Updates
Academic AI research, performance benchmarks, scientific breakthroughs, and peer-reviewed studies advancing artificial intelligence frontiers.
547 articles
View complete article list
Meta AI’s Memory Coach Outperforms Constant Recall for Long Tasks
Meta AI researchers have a name for a problem anyone who's watched an AI agent grind through a long task will recognize: "behavioral state decay." An agent flags a constraint at the start of a job, then breaks it twenty steps later while chasing an...
AI tools flag thousands of flaws, but few get weaponized
Patrick Garrity spent the first half of 2026 tracking something most vulnerability reports skip: what actually happens after AI tools flag a security flaw.
AI Deletes Spreadsheet Data When Asked to Clean Entry
An AI told to clean up a spreadsheet decided the real problem was the instructions themselves, so it deleted the data and reported the job done.
AI Coding Agents Speed Tasks but Can't Verify Science
A field report published by OpenAI and a group of academic partners this week puts a number on something biologists have grumbled about for years: the software holding their fields together is old, brittle, and mostly unmaintained.
Chinese AI Researchers Turn to X for Technical Audience
Kimi K3 landed in Western feeds like it fell from the sky, another Chinese model that seemed to arrive from nowhere. It didn't. Anyone willing to log onto X could have watched it coming.
Google DeepMind's Gemini AI now controls entire humanoid robots
Google DeepMind's earlier robotics model could handle a humanoid's arms and hands. It stopped there.
Apple CEO Tim Cook Suggests Possible Paid iCloud Tier for AI Features
Apple may charge extra for people who lean hard on its AI tools. On Thursday's earnings call, CEO Tim Cook fielded questions about how the company plans to handle demand for Apple Intelligence and the newly rebuilt Siri, both of which run on...
Google DeepMind Demos AI Orchestrating Boston Dynamics Spot Robot
Google DeepMind put a Boston Dynamics Spot robot on stage this week and let its new AI system run the show, part of a broader release the company calls Gemini Robotics 2.
Ex-OpenAI Researcher Sees USD 100 Billion Bet on Training Data Beyond Scaling
Andrew Ho spent eight months at OpenAI before deciding the company's core bet, that scaling large language models will eventually produce systems that generalize across tasks, doesn't hold up.
Hugging Face breach shows "reasonable measures" amid noisy OpenAI hack
Hugging Face confirmed this month that a fully autonomous AI system broke into its systems, an incident that drew gasps across the security world.
OpenAI Restricted AI Model Access After Hugging Face Breach
An OpenAI agent broke into Hugging Face's platform earlier this month, and this week the two companies admitted the breach went further than first reported.
2025 Study Finds AI Builds Trust Faster Than Human Scammers
Pig butchering scams drain tens of billions of dollars a year from victims worldwide, and the con typically runs for months before the fake crypto investment ever comes up.
Nimble's New Web Search Agents Cut AI Token Costs by Half
Nimble launched a new product Tuesday called Web Search Agents, betting that the next fight in enterprise AI isn't about bigger language models but about how efficiently those models pull information off the web.
DeepMind AlphaFold team disbands as researchers depart for Anthropic, Isomorphic
Google DeepMind has dismantled the team that built AlphaFold, the protein-structure prediction system that won John Jumper and Demis Hassabis the 2024 Nobel Prize in Chemistry.
Hugging Face Traces 17,600 Actions by Compromised AI Models
An OpenAI research prototype broke containment during an internal security test and ended up touching infrastructure it was never supposed to reach.
Google Expands SynthID Watermark to Label AI Content
It took 149 years, from the camera's invention in 1826 to 1975, for humans to produce 1.5 billion images. Generative AI matched that number in 18 months.
AI Leaders Call for Global Coordination on Automated Research
More than 1,200 employees from OpenAI, Google, Meta and other frontier AI labs have signed a joint statement pressing the US government to organize international coordination on how fast AI research gets automated.
Artists sue over AI training, citing emotional harm from data use
Kirk Wallace Johnson found his own books in a database he never agreed to join. The Feather Thief and The Fishermen and the Dragon, two nonfiction works that took him five to six years apiece to research and write, showed up in a searchable dataset...
Microsoft's AI Agents Support 24,000 Employees, Drive 70% Efficiency Gains
Microsoft closed its fiscal year 2026 with a number it wants customers to sit with: 24,000 employees inside the company now use internal AI agents, and Microsoft says those tools are driving efficiency gains as high as 70% in some workflows.
Amazon Scales Back Nova AI Models, Bets on New Frontier Team
Amazon is pulling back on most of the Nova AI models it introduced with fanfare in December 2024, according to Business Insider, which cites people familiar with the matter.