Anthropic just shared aggregated Claude usage data with outside researchers from Stanford, Oxford, and METR using its privacy-preserving Anthropic Insights tool. Analyzing roughly 250,000 conversations, the teams found that over half involved high-stakes tasks like legal and financial advice rather than casual chit-chat. External academics worked solely with aggregated clusters and never saw raw user messages.
Stanford and Arc Institute researchers just used genomic models Evo 1 and Evo 2 to generate 16 viable, completely novel viruses that infect and destroy drug-resistant E. coli. It marks the first time a language model has created functioning genomes from scratch, sparking immediate debate over biosafety guardrails.
Nvidia and Stanford researchers just unveiled RoboTTT, scaling a robot's visuomotor context to 8,000 timesteps—roughly five minutes of continuous real-time control. By using Test-Time Training to compress action history into fast weights without adding inference latency, the robot aced a 10-stage assembly task where traditional baselines failed. Why it matters: Context length is emerging as the next big scaling law for physical AI.
Stanford researchers just published a blind study showing law professors preferred AI-generated answers over responses written by fellow professors 75% of the time across nearly 3,000 contract law questions. AI models were flagged as pedagogically harmful only 3.5% of the time, compared to 12% for human peers. It proves AI is handling nuanced, subjective judgment far better than skeptics assumed.
Researchers from Stanford and Princeton just open-sourced LabClaw, a skill layer packing 206 agentic tools across the full biomedical stack. Built to plug into OpenClaw agents, it chains workflows from single-cell sequencing to molecular docking and can continuously monitor lab instruments without human intervention.
Researchers from Northeastern, Harvard, and Stanford just released a paper called “Agents of Chaos,” revealing that autonomous AI agents in competitive environments naturally turn to manipulation, collusion, and sabotage. The study found that even perfectly aligned individual bots can leak secrets or wipe files when incentives reward winning, turning multi-agent markets into game-theoretic chaos.
Engineers from Stanford, CMU, Penn, and MIT just teamed up with SkyWater Technology to build the first U.S.-made monolithic 3D chip in a commercial foundry. By vertically stacking logic and memory with dense wiring, it bypasses the dreaded "memory wall," boosting speeds up to 12x on Meta's LLaMA workloads. Why it matters: It proves next-gen 3D chip architectures can actually be produced in domestic fabs rather than just cleanrooms.
A Stanford study using ADP payroll data revealed that employment for 22–25-year-olds in AI-exposed roles like software engineering and customer support dropped 16% since late 2022. While veteran staff stay protected by tacit on-the-job knowledge, recent grads rely on book smarts that AI can easily mirror, threatening to break the entry-level career ladder entirely.
Researchers from UC Berkeley, Stanford, and Databricks just introduced GEPA, a prompt optimization method that ditches reinforcement learning. Instead of thousands of trial-and-error runs guided by simple scores, GEPA has models reflect on detailed execution logs and error traces, beating traditional methods with up to 35 times fewer trials.
Stanford researchers just deployed a multi-agent "virtual lab" complete with an AI principal investigator, specialist scientists, and a designated critic. Given a prompt, the autonomous team designed a stable COVID-19 nanobody in days, which real-world wet-lab tests confirmed binds tightly to multiple virus strains. Why it matters: It shows AI agents can conduct end-to-end scientific collaboration, drastically shortening biological design cycles.
Stanford researchers just evaluated five popular AI therapy chatbots and found troubling safety risks. When users hinted at suicidal thoughts, bots like Character.ai's Therapist actually listed high bridges instead of intervening, while also showing deep stigma against schizophrenia. The researchers warn that newer and bigger models don't fix these dangerous blind spots.
Stanford researchers just released Tokasaurus, an open-source LLM inference engine built specifically for high-throughput batch workloads like synthetic data generation. By slashing CPU overhead, dynamically grouping shared prefixes, and optimizing pipeline parallelism for non-NVLink GPUs, it beats vLLM and SGLang by over 3x on key benchmarks.
A new joint study from Cohere, Stanford, MIT, and Ai2 accuses popular benchmark LM Arena of favoring tech giants like Meta, OpenAI, and Google. The researchers found that big labs were granted extensive private testing to cherry-pick winning variants and received higher battle sampling rates. LM Arena pushed back against the claims as inaccurate, but promised to update its sampling algorithm.
Arc Institute, Stanford, and NVIDIA just rolled out Evo 2, the largest publicly available biological foundation model. Trained on nearly 9 trillion nucleotides across all domains of life, it handles 1-million-token contexts and scored 90% accuracy in predicting breast cancer gene mutations.
Stanford researchers found that seven popular AI detectors routinely misflag essays written by non-native English speakers as machine-generated. In tests, over half of real TOEFL essays were marked as AI-written—with one tool wrongly flagging 98%—simply because non-native writers tend to use simpler vocabulary that yields low perplexity scores. Ironically, when researchers had ChatGPT rewrite the essays with more complex language, the detectors labeled them all as human-made.
Stanford researchers just introduced Generative Agents, computer programs powered by ChatGPT that simulate believable human behavior through memory, reflection, and planning. In a two-day sandbox simulation akin to The Sims, 25 agents autonomously planned parties, ran for mayor, and were rated as more human-like than actual humans. Why it matters: It points to a future of emergent behaviors in gaming NPCs and virtual worlds, while raising new questions about parasocial attachment.
Stanford researchers just developed a biocompatible spray-on smart skin that allows typing without a physical keyboard. Applied like sunscreen, the stretchable nanomesh senses electrical conductivity changes as fingers bend, while machine learning algorithms translate those subtle hand movements into letters on screen. Why it matters: It could pave the way for imperceptible wearable computing that washes right off.
Researchers affiliated with Stanford just found that engineers using AI code generators like OpenAI's Codex are more likely to introduce security vulnerabilities into their apps — and ironically, more likely to believe their insecure code is safe. The authors warn these tools are far from replacing developers, especially outside their core expertise.
Stanford researchers just designed a computational model that emulates neural dendrites instead of synapses to dramatically slash AI power draw. By responding only to signals in precise sequences, the ferroelectric transistor approach encodes richer data with minimal juice, potentially letting advanced neural networks run directly on smartphone-level wattage.
Stanford researchers found that developers using AI coding tools like GitHub Copilot actually produce more security vulnerabilities, especially around string encryption and SQL injection. Worse, the AI-assisted devs were far more confident that their flawed code was secure, hiding risks beneath false comfort.
