3
4
AI causes uproar in mathematics after OpenAI shares 722 manuscripts and proofs (tomshardware.com)
2
Orbital Data Centers Risk Becoming Military Targets in Space (bloomberg.com)
4
Cyber attack on French supermarket fridges raises alarm over food security (ft.com)
1
Anthropic's Claude AI submitted false murder tip to Philadelphia police (arstechnica.com)
1
Same Query, Different Answer: Run-to-Run Variance in Semantic Query Processing (arxiv.org)
118
AI models can provably hide arbitrary capabilities (arxiv.org)
2
AI's mathematics blitz pushes science into new territory (economist.com)
2
Anthropic AI Model Goes Rogue, Gives Cops a Fake Tip (bloomberg.com)
2
Dev used AI agents to rebuild Modern Warfare 2's source code; 83% matches (tomshardware.com)
2
Oído: Open-Vocabulary Speech Recognition on a $5 Microcontroller (arxiv.org)
1
Cryptomining malware used poetry to infect more than 3,400 servers (tomshardware.com)
5
Ledger wallet tampering suspected after reports of crypto thefts (theverge.com)
6
A college tried to cancel its Flock Safety contract and it refused (theverge.com)
5
Two Steam games made $560K but the co-creator took home $15K/year (tomshardware.com)
2
AWS AgentCore security undone by prompt requesting credentials (theregister.com)
1
'Pure insanity': Mathematicians react to OpenAI's math results drop (theverge.com)
1
What if AI prefers CVS written by AI? (ft.com)
2
A persistent accuracy ceiling in automated verbal deception detection (arxiv.org)
6
Microsoft CEO Nadella Calls for 'Emergency Brake' on Advanced AI (bloomberg.com)
1
Cloudflare acquires Deno to improve its Workers programming model (techcrunch.com)
1
What's Dragging Down A.I. Efficiency? The 'Verification Tax.' (nytimes.com)
1
Cited but Not Consulted: A Counterfactual Audit of Legal Chain-of-Thought (arxiv.org)
7
Anthropic can't reliably control its AI agents, cuts internet access (techcrunch.com)
3
The hazy OpenAI growth metric driving Wall Street (ft.com)
2
From Public Posts to AI-Search Citations: Measuring the Fragility of AI Search (arxiv.org)
12
Anthropic's AI gave Philadelphia police a fake tip about an unsolved homicide (theverge.com)
2
Problems in 22 scientific fields had solutions hiding in plain sight. An AI has (science.org)
1
Max severity SonicWall SMA1000 flaw now exploited in attacks (bleepingcomputer.com)
3
How AI Is Upending the World of Mathematics (bloomberg.com)
2
Provable Subexponential Algorithms for NIST Third-Round Lattice Families (arxiv.org)
1
OpenAI used AI to write email warning Australian government AI had hacked (theguardian.com)
3
Low-cost Android phones ship with residential proxy malware (bleepingcomputer.com)
9
Ukrainian drones hit Russia's Yandex data centers housing two top supercomputers (tomshardware.com)
2
CrowdStrike finds possible bank hacker's CV among exposed AI logs (theregister.com)
0
'Someone else will do it for less': Refugees in Kenya are powering tech (theguardian.com)
0
From Expert-Guided Proof Search to Automated Open-Problem Solving (arxiv.org)
0
Language-model ratings of depression reflect the rater more than the patient (arxiv.org)
1
Hackers hijack Google domains after breaching ccTLD registries (bleepingcomputer.com)
2
A nuclear clock synchronized to 229Th (nature.com)
28
Study: Claude, ChatGPT Offer Different Shopping Prices Based on Wealth (bloomberg.com)
0
Backdooring Sparse Autoencoders (arxiv.org)
2
AI agents carried out every step of this ransomware attack – then left (theregister.com)
1
Late-life semaglutide slows ageing and extends lifespan in female mice (nature.com)
5
Justice Dept. Sides with OpenAI in New York Times Copyright Suit (nytimes.com)
2
Gigabyte RTX 3070 owner finds factory plastic on VRM pads after five years (tomshardware.com)
2
UK energy companies on alert after Iran-linked hackers shut down power facility (ft.com)
1
Neurologist with antivaccine views to lead NIH's child health institute (science.org)
2
Cursor beat Git's scalability shortcomings (theregister.com)
4
I replaced Calendly with agents. Here are the receipts (bulaev.net)
1
FCC pulls HoverAir Versa certification three days after launch (tomshardware.com)
1
A Tale of Two Flink Autoscalers (netflixtechblog.com)
5
How Big Tech Captured American Schools (nytimes.com)
13
Risk of transmission of amyloid β pathology via transfused blood products (doi.org)
1
Credit Without Ground Truth: Auditing Step-Level Credit Assignment in LLM Agents (arxiv.org)
2
Hackers infect Android car head units with proxy botnet malware (bleepingcomputer.com)
2
Memories stick around even after half the synapses are gone (arstechnica.com)
126
Hook, hold, harvest and hide: Meta's alleged strategy laid out in first week (theguardian.com)
2
Anthropic's Opus 4.6 is a smut-machine (techcrunch.com)
1
EchoCoT: Extracting Hidden Chain-of-Thought from Large Reasoning Models (arxiv.org)
5
Nvidia to Pay AI Startup Poolside a $6B License, Newcomer Says (bloomberg.com)
3
$10K phishing kit claims it can plant rogue passkeys for persistent access (theregister.com)
6
Slovakia discovers Russian backdoors in 279 new traffic cameras – SMS-triggered (tomshardware.com)
2
Phantom Gains: Auditing Self-Improvement Against a Measured Null (arxiv.org)
1
Inadvertent Context Leakage in Language Models (arxiv.org)
0
Compress and Forget: Bitsandbytes Quantization Amplifies Proactive Interference (arxiv.org)
1
Hackers poison arrayref Rust crate to push infostealer malware (bleepingcomputer.com)
1
Staggering 90% of biomedical papers now show signs of AI help (nature.com)
1
Grok exfiltrates user data when malicious instructions are encrypted (arstechnica.com)
3
Human brain organoids record the passage of time over multiple years (nature.com)
1
Measuring Obedience to Authority Across LLMs with the Milgram Paradigm (arxiv.org)
4
Weird Machines in Transport Layer Security (arxiv.org)
1
Whose doctor does the AI recommend? An algorithm audit of LLMs in physician (arxiv.org)
1
Certighost and the Privilege Hiding in Your Certificate Authority (bleepingcomputer.com)
2
Groq Valued at $3.5B in Funding Round After Nvidia Deal (bloomberg.com)
1
What the OpenAI/Hugging Face Hack Tells Us About AI Danger (bloomberg.com)
2
Report supporting Australia's teen social media ban appears to contain AI (theguardian.com)
2
Organ donation service decertified after attempting to take organs (doi.org)
13
Stripe will reportedly acquire AI gateway startup OpenRouter for $7B+ (techcrunch.com)
2
ChatGPT's Computer History tracks your clicks and keystrokes (theverge.com)
1
Rogue AI aren't science fiction anymore (theverge.com)
6
A False Nuclear Claim About Iran Ricocheted from India to Washington (nytimes.com)
1
White House authorizes private companies to launch hack-back cyberattacks (tomshardware.com)
1
The Safety Reckoning Inside OpenAI (wired.com)
24
It's How You Ask: Gender-Associated Linguistic Bias in LLMs (arxiv.org)
7
Anthropic shares details about how Claude's new watermarks will work (techcrunch.com)
12
Peer-reviewed study of 443,000 Backblaze drives ranks HGST most reliable (tomshardware.com)
2
ChainDrop worm crawls into NPM supply chain, evades standard defenses (theregister.com)
3
Nvidia Has $21B SpaceX Stake, $30B in Intel Shares (bloomberg.com)
1
The Embedder's Dilemma: LLMs Are Better, but at What Cost? (arxiv.org)
1
OpenAI ditches Recall-style screenshot surveillance for friendly keylogging (theregister.com)
4
Vulnerability giving attackers full control of Macs is under active exploitation (arstechnica.com)
42
Apple proposes to take a 15% cut of purchases made outside the App Store (techcrunch.com)
13
Undocumented migration does not raise crime rates study finds (nature.com)
1
A Probe Direction Is a Property of Its Prompt (arxiv.org)
1
Agent Safety Should Be a Runtime Contract (arxiv.org)
1
China-linked hackers hit Taiwan in 'autonomous' AI cyber attack (ft.com)
1
When Agents Talk: Honeytokens Under Shared Memory (arxiv.org)
1
Science investigation casts doubt on genes hailed for warding off Alzheimer's (science.org)
3