4
1
In-House LLM Serving at Netflix (netflixtechblog.medium.com)
3
New Windows LegacyHive zero-day gives hackers admin privileges (bleepingcomputer.com)
2
LLM Evaluators are Biased across Languages (arxiv.org)
11
Kalshi says it caught Trump's teleprompter operator insider trading (theverge.com)
5
Tesla driver who blamed crash on autopilot pressed accelerator 100%, NTSB finds (arstechnica.com)
2
Security incident disclosure – July 2026 (huggingface.co)
2
Tech support scam caused data breach at Australian airline Qantas (theregister.com)
2
The Refusal Residue: When Probes Catch Alignment Faking and When They Don't (arxiv.org)
8
Windows 0-day drops the same day Microsoft releases record number of patches (arstechnica.com)
2
Universal gates from braiding and fusing anyons on quantum hardware (nature.com)
2
Vint Cerf is working on a plan to unleash AI agents on the open internet (techcrunch.com)
2
Show HN: I spent a month turning LinkedIn into software (bulaev.net)
1
How Agentic Is Agentic Commerce? Measuring X402 Adoption and Authenticity (arxiv.org)
1
Skills That Don't Exist: A Large-Scale Study of Hallucinated Skill (arxiv.org)
2
Frame: A new X11 server – implemented directly in assembly (theregister.com)
2
Auditing the Risk Claims of Distributional Reinforcement Learning (arxiv.org)
1
The Unfair Judge: A Mechanistic Interpretability Account of LLM-as-Judge (arxiv.org)
1
Rethinking MCP Security: A Large-Scale Study of Runtime MCP Servers (arxiv.org)
2
Show HN: The New Cyborgs Don't Have Implants (bulaev.net)
2
Deceptive Grounding: Entity Attribution Failure in Clinical RAG (arxiv.org)
1
Seeing Is Free, Speaking Is Not: Uncovering the True Energy Bottleneck in Edge (arxiv.org)
2
Statistically Undetectable Backdoors in Deep Neural Networks (arxiv.org)
1
Failure as a Process: An Anatomy of CLI Coding Agent Trajectories (arxiv.org)
3
HTTP gets a QUERY method so complex searches can stop pretending to be POST (theregister.com)
2
Is the top song on Australian radio AI-generated? (theguardian.com)
4
AI rebrands fail to deliver a lasting share price boost (ft.com)
3
OpenAI Engineer's 'LOL' Moment Set Stage for Legal Fight with Apple (bloomberg.com)
2
RTX 5070 Ti throttles at 107C; Blackwell hotspot sensor readable via MODS tool (tomshardware.com)
9
New York Times says OpenAI hid evidence in ChatGPT copyright trial (techcrunch.com)
4
Nobel-winning chemist leaves US to direct AI materials lab in China (nature.com)
1
Large language models can predict the results of social science experiments (nature.com)
1
Beyond Refusal: Aligned vs. Abliterated LLMs for Vulnerability Analysis (arxiv.org)
5
British 'First Fleet' brought smallpox to Australia and may have killed millions (science.org)
3
GitHub AI agent leaks private repos when asked nicely (theregister.com)
2
Thermodynamic Measure of Intelligence (arxiv.org)
3
Show HN: Publora – One API/MCP for AI agents to post across 10 social networks (publora.com)
2
Steganography Without Modification: Hidden Communication via LLM Seeds (arxiv.org)
2
Attackers had month-long head start on patched Check Point VPN zero-day (theregister.com)
2
Waymo bought Apple's self-driving car proving ground for $220M (techcrunch.com)
1
CaaStle CEO Confessed to Fraud. The Board Let Her Stay (nytimes.com)
2
Creative Sound Blaster Katana V2X hijackable over Bluetooth; vendor denies flaw (tomshardware.com)
3
Huawei post-trained DeepSeek's 1.6T model on 1k Ascend 910C chips (tomshardware.com)
1
Dense Contexts Are Hard: Lexical Density Limits LLM Context Windows (arxiv.org)
1
OpenAI's Codex chained decade-old DoS attacks to crash web servers (theregister.com)
1
Same Weights, Different Robot: A Deployment Safety View of VLA Policies (arxiv.org)
1
Forge: Multi-Agent Graduated Exploitation and Detection Engineering (arxiv.org)
2
Spikes in LLMs Are Bias Vectors: Spike-Free Quantization (arxiv.org)
2
Forget Attention: Importance-Aware Attention Is All You Need (arxiv.org)
3
AI Investments Are 'Circular Bet' as ROI Disappoints, Bain Survey Shows (bloomberg.com)
77
LLMs Are Closer to Religion Than They Appear (theregister.com)
2
Too dangerous to release: is Mythos the start of the restricted-AI era? (nature.com)
8
Cox Media fined after bragging it spied on users through their phones (theverge.com)
1
Mec builds first High-NA EUV-fabricated quantum dot qubit (tomshardware.com)
3
Google has seriously leaned into AI enshittification lately (theregister.com)
3
Ghost CMS SQL injection flaw exploited in large-scale ClickFix campaign (bleepingcomputer.com)
1
Advanced Shader Delivery on RX 9070 XT: up to 95% faster load times (tomshardware.com)
1
To A.I. Executives, We're All Just 'Meat Computers' (nytimes.com)
2
Laravel Lang packages hijacked to deploy credential-stealing malware (bleepingcomputer.com)
6
Megalodon chums the waters in 5.5K+ GitHub repo poisonings (theregister.com)
14
Domain-Camouflaged Injection Attacks Evade Detection in Multi-Agent LLM Systems (arxiv.org)
4
A hacker group is poisoning open source code at an unprecedented scale (wired.com)
1
Is Capability a Liability? More Capable Language Models Make Worse Forecasts (arxiv.org)
22
GitHub faces a fight for its survival at Microsoft (theverge.com)
2
Imaging hidden objects with consumer Lidar via motion-induced sampling (nature.com)
1
Auditing Apple's DifferentialPrivacy.framework: Bugs, Misconfig, Practical Risks (arxiv.org)
2
Even Claude agrees: hole in its sandbox was real and dangerous (theregister.com)
4
SpaceX IPO Filing Reveals Anthropic Is Paying $15B/Year to Access Data Centers (wired.com)
2
Agent Security Is a Systems Problem (arxiv.org)
1
Max-severity flaw in ChromaDB for AI apps allows server hijacking (bleepingcomputer.com)
3
N display of stupid, secret CISA credentials found in public GitHub rep (arstechnica.com)
2
'It is incredible': How AI is transforming mathematics (nature.com)
1
Polynomial Quantum Attack on Module-LWE over Power-of-2 Cyclotomics (arxiv.org)
1
AI-Mediated Communication Can Steer Collective Opinion (arxiv.org)
2
Known by Their Actions: Fingerprinting LLM Browser Agents via UI Traces (arxiv.org)
1
Perfect Timing, Big Payouts: Insider Trading Red Flags Emerge on Polymarket (nytimes.com)
1
Oracle Poisoning: Corrupting Knowledge Graphs to Weaponise AI Agent Reasoning (arxiv.org)
42