Wired traced how AI embedded itself in the Iran conflict during active warfare. A company tied to an ICE detention facility just signed $132 million in contracts to house AI data center workers. The Pro-Human Declaration calls for a superintelligence ban and mandatory off-switches. MIT Technology Review found no clear law preventing military AI surveillance of Americans. Plus: OpenClaw 2026.3.8 adds backup and restore tooling, and dedicated OpenClaw hardware is now a category.

Hundreds of experts signed the Pro-Human Declaration calling for a superintelligence ban. MIT Tech Review found no law actually prevents military AI surveillance of Americans. Wired traced how AI companies embedded themselves in the Iran conflict. AWS launched a HIPAA-eligible AI agent platform for healthcare at $99/month per clinician. Plus: OpenClaw 2026.3.7 ships a ContextEngine plugin interface, and Google's Workspace CLI makes Gmail and Drive accessible to AI agents.

A startup built an orb to jam AI microphones — physics says it probably won't work. City Detect is mounting AI cameras on garbage trucks to flag code violations. Netflix acquired Ben Affleck's filmmaking AI company for production efficiency, not synthetic content. The ChatGPT backlash is growing as Claude downloads surge. Apple's biggest Mac refresh in a year runs from $599 to $3,899. Plus: Qwen 3.5 small models outperform OpenAI's 120B on laptops, and the chardet GPL relicensing controversy.

Open source maintainers now have a protocol for AI-generated PR spam. Meta's smart glasses privacy story escalates with worker accounts of annotating intimate user data. GPT-5.4 launches in Microsoft Foundry with a pitch around production reliability and verifiable reasoning. ChatGPT's default behavior became a meme, and OpenAI published chain-of-thought controllability research with built-in cyber mitigations.

Qwen 3.5 drops a practical fine-tuning guide, Nvidia doubles down on open models at GTC, and the Alibaba/Qwen trust wobble shows how fast developer confidence can shift. Plus AMD brings Ryzen AI to mainstream desktops, a hard safety story around Google's AI and mental health harm allegations, and an open source roundup on OpenClaw on Amazon Lightsail and Nvidia's telco model push.

Anthropic's Claude crashed right when everyone wanted to use it, Apple built a billion-dollar server farm that nobody needs yet, and the QuitGPT boycott just passed 1.5 million participants. Meanwhile, Sonnet 4.6 told users it was DeepSeek-V3, and the open source AI agent ecosystem is wrestling w...

AI coding tools are creating a generation of developers who can ship fast but can't explain why. OpenAI's Sam Altman admitted the Pentagon deal was rushed and the optics are bad. Google's SynthID watermark just caught a fake photo of Khamenei's body. AMD proved you can run a trillion-parameter mo...

San Franciscans turned the sidewalk outside Anthropic's office into a chalk memorial after the Pentagon blacklist; Burger King deployed an AI politeness cop for 500 locations; the real cost of AI coding tools is roughly nine times the sticker price; NanoClaw published a security teardown arguing you shouldn't trust any AI agent; and Anthropic's CEO told the Pentagon in writing they cannot hand over unrestricted access to Claude.

Pentagon designates Anthropic a supply-chain national security risk after refusing autonomous weapons and surveillance access; Sam Altman publicly defends his rival; Perplexity launches Computer, a 19-model super agent; Anthropic publishes the AI Fluency Index showing 85% of users iterate on responses but scrutiny drops for polished outputs; and just 4% of companies have achieved ROI on AI investments.

Anthropic rejects the Pentagon's final offer on AI safeguards, refusing to remove guardrails for surveillance and autonomous weapons; Goldman Sachs says AI investment added basically zero to US economic growth in 2025; Google launches Nano Banana 2 image model with Pro-quality results at Flash speed; Ladybird browser uses AI agents to port 25,000 lines of C++ to Rust in two weeks; and the agent economy raises an unanswered question — when AI signs an SLA and doesn't deliver, who settles it?