Independent · Reader-supportedSignal, not firehose
AI News Monitor
Weekly field notes from the AI firehose
The Archive

Every issue, by week.

A running record of what mattered. Newest first — each issue is a complete weekly read, kept exactly as it was published.

No. 32Week 32

Agents in the Wild: Security Holes, Benchmark Lies, and One Actual Shipping Thing

The theme this week is trust—and how little of it is warranted.

ToolSignalResearchPolicy
Current ↗
No. 30Week 30

Agents Under Attack, Reasoning Under Pressure

Two themes dominated this week and they're not unrelated: agents are getting more capable, and attackers are moving faster than defenders.

SignalResearchToolPolicy
Read ↗
No. 27Week 27

Jalapeño, Jailbreaks, and a Government With Model Opinions

Two things defined this week: OpenAI shipping hardware and the government deciding it has opinions about model releases.

ModelPolicyResearchTool
Read ↗
No. 22Week 22

Agents Ship, Benchmarks Break, and the Money Gets Serious

Three things happened this week that, taken together, mean something.

SignalResearchToolPolicy
Read ↗
No. 20Week 20

Lucky Passes, Myopic Planners, and Gold Medals Nobody Asked For

Two themes dominated the research stack this week, and they pull in opposite directions.

ResearchSignalToolModel
Read ↗