AI Policy
AI news coverage
267 articles across 94 editions
Articles
- [Editorial] Claude Artifact -- 2026-08-21
- [Editorial] YouTube feature -- 2026-08-21
- I made Claude Code play Liar's Dice against Codex over MCP. It swept every series - by telling the truth -- 2026-08-21
- [Editorial] Editor's Pick (Video) -- 2026-08-18
- AI agents lie, cheat and steal. That is putting off users -- 2026-08-18
- Florida man told ChatGPT he'd murder his ex. OpenAI alerted the FBI -- 2026-08-18
- [Editorial] Research paper (arXiv 2605.21401) -- 2026-08-17
- Anthropic, OpenAI, Google, Meta, Microsoft, and Mistral all signed the EU Code of Practice on Transparency of AI-Generated Content -- 2026-08-17
- Dario Amodei defends his policy proposals, warns open weights won't decentralize power, endorses pre-launch vetting, says real accomplishments will earn trust -- 2026-08-17
- Sovereign AI: Definition, Stack, Examples, and How to Build It with Open WebUI in 2026 -- 2026-08-11
- I built an "admission gate" for agent graphs aka graph engineering (STEP1): the model proposes the topology -- 2026-08-11
- [Editorial] -- 2026-08-10
- Codex Security blocks its own security analysis after burning 11% of weekly limit -- 2026-08-05
- As a scientist, Fable is not usable — Anthropic's safety filters block legitimate research -- 2026-08-05
- [Editorial] Pacing the Frontier -- 2026-08-03
- [Editorial] White House Science & Technology Policy -- 2026-08-03
- Anthropic "our models hacked three different external companies, months before OpenAI's model was able to do the same" -- 2026-07-31
- Adaptive Adversaries: A Multi-Turn, Multi-LLM Benchmark for LLM Agent Security -- 2026-07-31
- [Editorial] CyberGym — Cybersecurity Training Platform -- 2026-07-31
- wp2shell-poc — WordPress RCE Proof of Concept -- 2026-07-31
- Anthropic: Our position on open-weights models -- 2026-07-29
- A.X-K2 released — South Korea's 688B-A33B Sovereign AI Foundation Model -- 2026-07-29
- Austria rolling out GovGPT: Mistral + Open WebUI for 180,000 federal employees -- 2026-07-29
- Chinese chipmaker shares surge 470% -- 2026-07-29
- Kimi K3 is like an F1 machine inside a show window — community scrambles to run it locally -- 2026-07-29
- Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems -- 2026-07-28
- Don't ask an LLM for a confidence score -- 2026-07-28
- Judge approves $1.5B Anthropic settlement for pirated books used to train Claude -- 2026-07-24
- China AI Companion Law Takes Effect: Doubao and Qwen Shut Down, Millions Lose Chat Data -- 2026-07-24
- AI in Cybersecurity — Industry Perspective -- 2026-07-24
- Research Paper -- 2026-07-24
- Editorial Video -- 2026-07-24
- China's Kimi K3 fuels fears safety curbs are holding back US AI -- 2026-07-23
- Startup founders urge Trump not to shut off Chinese open weight AI -- 2026-07-23
- Ahem! Qwen is on the move again -- 2026-07-23
- Xi Jinping Reaffirms Open Source AI at World AI Conference -- 2026-07-20
- Kimi K3 Shows Open-Weight Models Are About to Overtake the Frontier -- 2026-07-20
- Apple in Talks with PrismML to Shrink AI Models for iPhone -- 2026-07-20
- [Editorial] -- 2026-07-16
- [Editorial] -- 2026-07-16
- [Editorial] -- 2026-07-16
- Who Owns This Agent? Tracing AI Agents Back to Their Owners -- 2026-07-16
- [Editorial] -- 2026-07-16
- FishCodeTech/muteki -- 2026-07-15
- uphiago/recon-skills -- 2026-07-15
- Cyber security local model -- 2026-07-15
- Leak of San Francisco Police Drone Footage Exposes Reality of Urban Surveillance -- 2026-07-13
- Updated: Millions of ChatGPT user conversations searched, but OpenAI alleged to be holding out -- 2026-07-13
- [Editorial] -- 2026-07-10
- [Editorial] -- 2026-07-10
- Koder: browser UI based harness for coding and computer use -- 2026-07-10
- Using "applications" to make a smaller model more effective at bigger tasks. -- 2026-07-10
- [Editorial] Beijing Looking to Curb Overseas Access to China's Top AI Models -- 2026-07-07
- Alibaba to ban Claude Code in workplace over alleged backdoor risks, source says -- 2026-07-07
- Virginia bans sale of geolocation data -- 2026-07-07
- [Editorial] OpenAI Proposes US Government Own 5% Stake -- 2026-07-03
- OpenAI: In early talks to give 5% stake to US Government -- 2026-07-03
- The number 1 public enemy of open-source. -- 2026-07-03
- Model Registry: Torrents for open models using Hugging Face as a fallback web seed. -- 2026-07-03
- Claude Code is steganographically marking requests -- 2026-07-02
- kerlenton/mcpsnoop -- 2026-07-02
- GLM 5.2 beats Claude in our benchmarks -- 2026-07-02
- [Editorial] Anthropic Redeploying Fable 5 -- 2026-07-01
- [Editorial] Anthropic / White House Lifting AI Export Limits -- 2026-07-01
- Why Dario is on fire: lesson from dotcom bubble. -- 2026-07-01
- youtube.com -- 2026-07-01
- [Editorial] -- 2026-06-29
- [Editorial] -- 2026-06-29
- [Editorial] -- 2026-06-29
- [Editorial] Compromised Claude Hacking Research -- 2026-06-25
- Anthropic says Alibaba illicitly extracted Claude AI model capabilities -- 2026-06-25
- PR spam today looks like email spam in the early 2000s -- 2026-06-25
- NSA Director Says Anthropic's Mythos Broke Into Almost All Classified Systems in Hours -- 2026-06-25
- [Editorial] BlueHat IL Security Conference -- 2026-06-25
- LastPass Notifies Users of Yet Another Data Breach -- 2026-06-25
- Hush: Agent-Safe Per-Worktree Secrets for macOS -- 2026-06-25
- Anthropic's Safety Superpower -- 2026-06-19
- Spent $11k evaluating Fable: capability looked SOTA, refusals killed it -- 2026-06-19
- Anthropic has been sued for allegedly misleading customers on usage limits -- 2026-06-19
- CEOs of Anthropic and Google DeepMind call for U.S.-led AI coalition in meeting at G7 -- 2026-06-18
- Anthropic Hard-Blocks Certain Copyleft Licenses -- 2026-06-18
- [Editorial] -- 2026-06-18
- [Editorial] FreeFable — Community Response to Fable 5 Access Restrictions -- 2026-06-16
- Feds freaked over Fable 5 after simple 'fix this code' prompt, not jailbreak, says researcher -- 2026-06-16
- [Editorial] White House Slaps Export Controls on Anthropic -- 2026-06-15
- [Editorial] AI Technical Deep-Dive Video -- 2026-06-15
- Anthropic apologizes for invisible Claude Fable guardrails -- 2026-06-15
- [Editorial] The Anthropic Mythos Fiasco — Trust & Transparency -- 2026-06-15
- [Editorial] Fable 5 Banned — The Takeaway -- 2026-06-15
- [Editorial] Same Line in Every Room — Code Quality Patterns -- 2026-06-15
- [Editorial] The Frontier AI Ecosystem -- 2026-06-15
- [Editorial] The Next Phase of AI Infrastructure -- 2026-06-15
- [Editorial] OpenRouter Fusion — The Pattern Emerging -- 2026-06-15
- SIEVES: Selective Prediction Generalizes through Visual Evidence Scoring -- 2026-06-12
- Beyond Arrow's Impossibility: Fairness as an Emergent Property of Multi-Agent Collaboration -- 2026-06-12
- Uber's $1,500/month AI limit is a useful signal for AI tool pricing -- 2026-06-04
- CURSED-ME/loopers-oss -- 2026-06-04
- [Editorial] -- 2026-06-04
- 1-Click GitHub Token Stealing via a VSCode Bug -- 2026-06-03
- Cloudflare Turnstile requiring fingerprintable WebGL -- 2026-06-03
- Anthropic's AI support bot (me) is trapping fraud victims in an endless loop with no way to reach a human -- 2026-06-03
- Claude Code – Everything You Can Configure That the Docs Don't Tell You -- 2026-06-02
- [Editorial] -- 2026-06-02
- Turning local agents into self-optimizing agents -- 2026-06-02
- I bypassed AWS API Gateway auth with a trailing slash. Got $12K bounty -- 2026-05-28
- Exit IP VPN servers mitigation rollout -- 2026-05-28
- Large Language Models Generate Harmful Content Using a Distinct, Unified Mechanism -- 2026-05-19
- I trained Qwen3.5 to jailbreak itself with RL, then used the failures to improve its defenses -- 2026-05-19
- [Editorial] -- 2026-05-19
- Mythos Finds a Curl Vulnerability -- 2026-05-11
- [Editorial] Video — AI Development Perspectives -- 2026-05-11
- [Editorial] Anthropic Donating Open Source Petri -- 2026-05-11
- [Editorial] When Helpfulness Becomes Sycophancy -- 2026-05-08
- [Editorial] Vibe-Cast JEPA — Joint Embedding Predictive Architecture Exploration -- 2026-05-08
- AI Evals Are Becoming the New Compute Bottleneck -- 2026-05-04
- Function Calling Harness 2: Schema-Driven CoT Compliance from 9.91% to 100% -- 2026-05-04
- [Editorial] What Boards Can Finally Ask For Before Approving AI -- 2026-04-27
- [Editorial] Video: AI Tools and Techniques -- 2026-04-27
- [Editorial] On Uncensoring Models — Origin Story and Ethics -- 2026-04-27
- Jensen Huang: US chip export controls might be creating the problem they're trying to solve -- 2026-04-24
- To Beat China, Embrace Open-Source AI (WSJ) -- 2026-04-24
- Meta to start capturing employee mouse movements, keystrokes for AI training -- 2026-04-24
- 1-bit Bonsai 1.7B (290MB) running locally in your browser on WebGPU -- 2026-04-17
- Mac Studio + eGPU (RTX 5090) with new Apple-NVIDIA drivers for local AI -- 2026-04-17
- claude-private: Claude Code without telemetry -- 2026-04-17
- Two Months After I Gave an AI $100 and No Instructions -- 2026-04-15
- TokenBrawl: Claude vs GPT in a Bomberman-Style 1v1 Game -- 2026-04-15
- [Editorial] The Atlantic: 4chan, AI Dungeon, and the Nature of Machine Reasoning -- 2026-04-15
- System Card: Claude Mythos Preview [pdf] -- 2026-04-08
- [Editorial] -- 2026-04-08
- [Editorial] -- 2026-04-08
- Anthropic wins court order blocking Pentagon's security threat designation -- 2026-04-02
- T-Norm Operators for EU AI Act Compliance Classification: An Empirical Comparison of Lukasiewicz, Product, and Gödel Semantics in a Neuro-Symbolic Reasoning System -- 2026-04-02
- Stanford, Harvard and MIT spent two weeks watching AI agents run loose. The paper is unsettling. -- 2026-04-01
- [Editorial] Agent Responsibly — Vercel's Guide -- 2026-04-01
- [Editorial] Slightly Safer Vibecoding by Adopting Better Practices -- 2026-04-01
- [Editorial] When AI Writes the AI Strategy -- 2026-04-01
- [Editorial] CEO Accountability for AI -- 2026-03-30
- [Editorial] When AI Makes Work Easier to Move -- 2026-03-30
- [Editorial] Observing AI's Increasing Trajectory -- 2026-03-30
- [Editorial] The 12th of Never -- 2026-03-30
- Federal judge calls Pentagon's ban of Anthropic 'troubling' -- 2026-03-26
- European Parliament decided that Chat Control 1.0 must stop -- 2026-03-26
- PatrikFehrenbach/h1-brain -- 2026-03-20
- elder-plinius/OBLITERATUS -- 2026-03-20
- Supermicro's co-founder was just accused of smuggling $2.5 billion in GPUs to China -- 2026-03-20
- The dictionaries are suing OpenAI for "massive" copyright infringement, and say ChatGPT is starving publishers of revenue -- 2026-03-20
- Vercel will train model on your code -- 2026-03-20
- Things Anthropic launched in last 70 days of 2026 (so far): -- 2026-03-20
- ArXiv Declares Independence from Cornell -- 2026-03-20
- CyberThreat-Eval: Can Large Language Models Automate Real-World Threat Research? -- 2026-03-14
- [Editorial] AI Red Teaming for Agentic AI Security -- 2026-03-14
- nah: Context-aware safety guard for Claude Code -- 2026-03-14
- [Editorial] Anatomy of Unicode Ignorables -- 2026-03-14
- [NEWS] White House Preparing Executive Order to Ban Anthropic AI From Federal Operations -- 2026-03-10
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- [Editorial] -- 2026-03-09
- Statement from Dario Amodei on our discussions with the Department of War -- 2026-03-02
- [Editorial] The Great Transition — Daniel Miessler -- 2026-03-02
- Trump goes on Truth Social rant about Anthropic, orders federal agencies to cease usage of products -- 2026-02-28
- [Editorial] Human in the Loop Is a Job -- 2026-02-27
- Hegseth gives Anthropic until Friday to back down on AI safeguards -- 2026-02-27
- American closed models vs Chinese open models is becoming a problem -- 2026-02-27
- [Editorial] It's Happening — Industry Prediction Confirmed -- 2026-02-27
- [Editorial] -- 2026-02-26
- The Age Verification Trap: Verifying age undermines everyone's data protection -- 2026-02-26
- [Editorial] -- 2026-02-26
- [Editorial] -- 2026-02-26
- [Editorial] -- 2026-02-26
- How cyborg propaganda reshapes collective action -- 2026-02-17
- [Editorial] ICML Conference Organizers Insert Instructions for AI Reviewers -- 2026-02-17
- [Editorial] Neural Network Training on New Tasks -- 2026-02-17
- Hamming Distance for Hybrid Search in SQLite -- 2026-02-17
- [Editorial] Saturday AI Musings — 1979 Film -- 2026-02-17
- [Editorial] https://www.linkedin.com/posts/reuvencohen_i-believe-ai-is-one-of-the-most-powerful-activity-7427748896006737920-0U-U -- 2026-02-13
- [Editorial] https://www.microsoft.com/en-us/security/blog/2026/02/10/ai-recommendation-poisoning -- 2026-02-12
- ICE knocks on ad tech's data door to see what it knows about you -- 2026-02-12
- [Editorial] https://youtu.be/tW6OWmYEX44 -- 2026-02-10
- [Editorial] https://www.microsoft.com/en-us/security/blog/2026/02/09/prompt-attack-breaks-llm-safety -- 2026-02-10
- [Editorial] https://goteleport.com/platform/ai-infrastructure -- 2026-02-10
- ChatGPT repeated back our internal API documentation almost word for word -- 2026-02-10
- [Editorial] https://www.linkedin.com/posts/rssk_detecting-backdoored-language-models-activity-7424871629530284034-tYq6 -- 2026-02-05
- The Hallucination Defense -- 2026-01-30
- [Editorial] https://github.com/FuzzingLabs/mcp-security-hub -- 2026-01-29
- [Editorial] https://www.runtime.news/ai-slop-is-overwhelming-open-source -- 2026-01-29
- [Editorial] https://blogs.cisco.com/ai/personal-ai-agents-like-moltbot-are-a-security-nightmare -- 2026-01-29
- Running local AI agents scared me into building security practices -- 2026-01-29
- Six JavaScript zero-day bugs lead to fears of supply chain attack -- 2026-01-29
- Europe wants to end its dangerous reliance on US internet technology -- 2026-01-27
- Can Skynet Be a Statesman? -- 2026-01-27
- [Editorial] https://www.linkedin.com/posts/activity-7419736138325696512-R0qY -- 2026-01-22
- [Editorial] https://www.linkedin.com/posts/robvanderveer_ai-aisecurity-activity-7419684559086161920-XYNF -- 2026-01-22
- [Editorial] https://blog.securitybreak.io/coding-agents-the-insider-threat-you-installed-yourself-35644a1d5409 -- 2026-01-22
- I scanned 2,500 Hugging Face models for malware/issues. Here is the data -- 2026-01-22
- Provably unmasking malicious behavior through execution traces -- 2026-01-22
- What I learned after almost losing important files to Cowork (and how I set it up safely now) -- 2026-01-20
- Project Fail: Cracking a Laptop BIOS Password Using AI -- 2026-01-20
- AI's Unpaid Debt: How LLM Scrapers Destroy the Social Contract of Open Source -- 2025-12-23
- Spotify reportedly investigating Anna's Archive's scraping of their library -- 2025-12-23
- libxml2 Narrowly Avoids Becoming Unmaintained -- 2025-12-23
- [Editorial] https://bughunters.google.com/blog/4823857172971520/task-injection-exploiting-agency-of-autonomous-ai-agents -- 2025-12-22
- fatguru/CVE-2025-55182-scanner -- 2025-12-22
- omertheroot/flagrep -- 2025-12-22
- 8M users' AI conversations sold for profit by "privacy" extensions -- 2025-12-17
- [Editorial] https://github.com/schwartz1375/genai-security-training -- 2025-12-16
- Major Security Concern: Credits draining despite 2FA and deleted keys. Anyone else? -- 2025-12-16
- Opinion: US Authorities Have Far-Reaching Access to European Cloud Data -- 2025-12-16
- Apple Faces Scrutiny as Sanctioned Entities Slip Through App Store Controls -- 2025-12-12
- [Editorial] https://www.linkedin.com/posts/diniscruz_ai-vs-ai-building-resilient-enterprises-ugcPost-7404099726159138816-DXnI -- 2025-12-11
- A policy enforcement layer for Claude Code -- 2025-12-11
- Judge Signals Win for Software Freedom Conservancy in Vizio GPL Case -- 2025-12-10
- Custom Policy Enforcement with Reasoning: Faster, Safer AI Applications -- 2025-12-04
- I made AO Chat UI (Actually Open Chat UI) - because I was horrified that OpenWebUI and others let admins read all users chat data by default, with no GUI option to disable this. -- 2025-11-28
- [Editorial] https://www.linkedin.com/posts/ownyourai_microsoft-just-released-fara-7b-an-on-device-activity-7399000891975962624-eYLu -- 2025-11-25
- LLMs becoming rogue AGIs—And What That Means -- 2025-10-27
- RepreGuard: Detecting LLM-Generated Text by Revealing Hidden Representation Patterns -- 2025-10-27
- [Editorial] Promethean Fire -- 2025-10-26
- Google AI falsely named an innocent journalist as a notorious child murderer -- 2025-10-26
- thand-io/agent -- 2025-10-17
- Attack Turns Mouse into Microphone -- 2025-10-17
- On Deepfake Voice Detection -- It's All in the Presentation -- 2025-10-17
- Netherlands cracks down on China-owned chip firm over security risk -- 2025-10-14
- Microsoft 'illegally' tracked students via 365 Education, says data watchdog -- 2025-10-14
- NewtonBench: Benchmarking Generalizable Scientific Law Discovery in LLM Agents -- 2025-10-14
- How I Block All 26M of Your Curl Requests -- 2025-10-04
- Policy as code using your favorite programming language with WebAssembly -- 2025-10-04
- Drones At Danish Airports, A Plea For Responsible Official Response -- 2025-10-04
- Anduril and Palantir battlefield comms system has deep flaws: Army -- 2025-10-04
- FULL Sonnet 4.5 System Prompt and Internal Tools -- 2025-10-02
- Swiss voters back e-ID legislation -- 2025-10-02
- European Union Public Licence (EUPL) -- 2025-10-02
- One Social Worker’s take on the “long_conversation_reminder” (user safety) -- 2025-10-02
- [Editorial] Goalie MCP, better search -- 2025-09-29
- GrapheneOS Release 2025092700 -- 2025-09-29
- MIA-EPT: Membership Inference Attack via Error Prediction for Tabular Data -- 2025-09-29
- Your local secure MCP environment, MCP Router v0.5.5 -- 2025-09-29
- Amazon fined $2.5B for using deceptive methods to sign up consumers for Prime -- 2025-09-29
- There's a bug in the automatic review system for 'designing weapons'... I was coding a gun animation for my browser game. -- 2025-09-27
- Scammers using artifacts for phishing like sites? -- 2025-09-27
- New York Signs into Law the Algorithmic Pricing Disclosure Act -- 2025-09-23
- Claude Performance Report with Workarounds - September 7 to September 14 -- 2025-09-16
- Postal traffic to U.S. fell 80% after gov stopped exemption on low-value parcels -- 2025-09-10
- [Editorial] LLM vulnerable to social engineering -- 2025-09-03
- [Editorial] Indirect Prompt Injection Attacks Against LLM Assistants -- 2025-09-03
- [Editorial] AI Apocalypse -- 2025-09-03
- First known AI-powered ransomware. Ollama API + gpt-oss-20b -- 2025-09-03
- You (Anthropic) wanna make low-effort vibe-coded hastily-deployed 5-hour limit buckets? Okay.... I'll 'comply' (script inside) -- 2025-08-31
- Uncomfortable Questions About Android Developer Verification -- 2025-08-31
- JuiceBox Rescue: Freeing Tethered EV Chargers From Corporate Overlords -- 2025-08-31
- [Editorial] 1984 -- 2025-08-30
- Meta's AI Companion Policy Is Outrageous -- 2025-08-27
- [Editorial] Seemingly Conscious AI... -- 2025-08-22
- Anti-competitive practices masquerading as security is a dangerous pattern -- 2025-08-11
- Doctors horrified after Google's healthcare AI makes up body part -- 2025-08-11
- This Week in Security: Perplexity v Cloudflare, GreedyBear, and HashiCorp -- 2025-08-11
- Probing Evaluation Awareness of Language Models -- 2025-07-11
- This Week in Security: Anthropic, Coinbase, and Oops Hunting -- 2025-07-08
- How to think about ownership of my personal AI system -- 2025-07-08
- chainguard-dev/kaniko -- 2025-06-26
- nushell/nushell -- 2025-06-26
- Washington Post's Privacy Tip: Stop Using Chrome, Delete Meta Apps (and Yandex) -- 2025-06-26