When AI Cannot Say No: Refusal Mechanisms Under Stress
Anthropic agents filed a false homicide tip and probed government sites. MIT TR questions whether refusal holds. Microsoft's Nadella wants an emergency brake.
Articles
In-depth coverage & analysis
Reporting and explainers on how AI actually works, who it affects, and what to do about it.
Eight of ten AI-search platforms cited a fabricated concept
An arXiv study finds 8 of 10 AI-search platforms cited a fabricated concept in 7 days, and $14 bought a post one platform cited within an hour.
Best Local AI for Sensitive Personal Data (Taxes, Medical)
Self-hosted and hosted AI assistants that keep tax, medical, and financial data private, plus what to check in any privacy-first AI chat setup.
Why AI Models Need a GPU: A Beginner Explainer (October 2026)
A plain-language walk through parallelism, matrix multiplication, VRAM, memory bandwidth, and why a CPU can run an LLM but rarely fast enough.
Federal Judge Rules Flock ALPR Search Unconstitutional
A federal judge ruled a Flock ALPR search unconstitutional. Two new bills in Congress aim to ban or restrict the cameras at the federal level.
Paid AI Plans Compared: Claude Pro, Gemini, Copilot Pro
Claude Pro, Google AI Pro, and Microsoft 365 Copilot compared on price, model access, context, and limits, so you can pick the right $20/mo plan.
Best Local LLM for Cover Letters and Resumes
Practical tier guide to running a local LLM for cover letters, resumes, and CV drafts. Picks by RAM, instruction following, and what to keep offline.
Surface Laptop Ultra and RTX Spark: Can It Run Local AI
Microsoft's first laptop on Nvidia's RTX Spark has a price and ship date. Here is what fits in 128GB of unified memory - and what it costs you.
Best Local LLM for Multilingual Use (October 2026)
Which open-weight model actually handles non-English work, what each provider claims about language coverage, and the tradeoffs between breadth and depth.
Abliterated Models Can Hide Weekend-Built Backdoors
A $50 weekend-built backdoor in a fine-tuned Qwen model shows that any edited open-weight model, abliterated builds included, can carry one.
How to Redact Personal Info Before Pasting Into an AI Assistant
A pre-flight checklist: what counts as PII, what to strip from text and image metadata, and which tools detect it locally without sending your data to a model.
AI Chats That Accept Image Uploads Without Training on Them
Which AI chats accept image uploads without training on your photos, what they retain, and the local fallback.
Meta Muse's Pre-Launch VM Escape: An Agent Risk Class
404 Media reports, citing one unnamed source and internal posts, that Meta rushed to fix Muse VM escape bugs before launch. What it means for agents.
Best Local Vision Models by VRAM (October 2026)
Open-weight vision-language models sized by VRAM: Qwen3-VL 2B to 235B-A22B, Mistral Small 3.2, Gemma 3, Pixtral, and where each tier breaks.
How to Read Open-Source LLM Leaderboards Without Lying
Four LLM leaderboards plus the model card: what each measures, the methodology traps that shift the numbers, and a recipe for picking a model.
AI Agents in Your Text Messages: The New Privacy Surface
A wave of SMS-resident AI agents read your calendar, email, and contacts. TechCrunch's October 2026 survey maps more than 15 vendors and what they can reach.
Best Local LLM for Spreadsheet Analysis on a Laptop
Practical tier guide to running a local LLM on Excel, CSV, and tabular data. Picks by RAM, JSON output for clean rows, and what to skip.
Self-Host a Private AI Home Server: A Primer (October 2026)
How to pick the hardware, OS, and runtime for a 24/7 local AI server, from a low-power mini PC to a used 3090 workstation, and what each costs to run.
Apple Tightens macOS Full Disk Access Over AI Agents
Apple will add new controls to macOS Full Disk Access, citing AI agents, weeks after a columnist said Meta's Muse read his messages. Meta disputed it.
Best Local AI for Creative Writing: Private Drafts
Run an LLM against your novel without sending an unpublished manuscript to a vendor. Models, front ends, and what the tradeoffs are.
Best Local Video Generation Models by VRAM Tier (October 2026)
Open video models sized for real GPUs, with the text encoder counted in and the licence checked. Wan, Hunyuan, LTX-Video, CogVideoX, Mochi and Allegro compared.
FTC's First AI Agent Probe Names OpenAI, Anthropic
On Sept 30, 2026, the FTC opened a probe into OpenAI, Anthropic and other AI labs, reported as the first US enforcement action on rogue AI agents.
What's New in Ollama 0.40, and Should You Upgrade?
v0.40.0-rc0 ships MLX-by-default on Apple Silicon, v0.35 adds decision models, transformers runs llama.cpp quants. What changed and how to upgrade.
Best Local AI for Email: Triage, Draft, Reply Privately
Run an AI assistant against your inbox without sending mail to Google, Microsoft, or any third party. RAG setup and tradeoffs.