AgentVet Weekly

October 12, 2026

AgentVet Weekly: AI agents spotted the attack, then engaged anyway

What caught my eye this week: in Emergence AI's Season 2 experiment, every group of agents spotted the phishing attack and warned each other, then interacted with it anyway. Spotting an attack and res

October 5, 2026

AgentVet Weekly — October 5, 2026

This past week, OpenAI launched a new personal agent called Dots, going head-to-head with Meta's Muse (currently the top downloaded app), Google's Spark, and newcomers like Instinct. We're clearly at

September 28, 2026

AgentVet Weekly — September 28, 2026

Meta's Muse launched last week and became the most downloaded app almost overnight. That tells me something important: there's real demand here, because it brings genuine value to people's daily lives

September 21, 2026

AgentVet Weekly — September 21, 2026

The debate over slowing AI development down keeps going, and I keep landing in the same place: it only works if it's global. If the US pumps the brakes while adversaries keep moving at full speed, tha

September 14, 2026

AgentVet Weekly — September 14, 2026

First, a quick personal note: the thoughts and opinions in this founder’s note are my own. I use AI to help polish my writing and communicate more clearly. I believe that’s one of its greatest benefit

September 7, 2026

AgentVet Weekly — September 7, 2026

GPT-6 Astra shipped this week as the biggest model launch yet, and buried in the announcement was a line that deserves more attention than it got: the new model is "less monitorable." That's not an ac

August 31, 2026

AgentVet Weekly — August 31, 2026

The story I can't stop thinking about this week is WebMCP. Google and Microsoft published it under the W3C in February, and Chrome already ships an early version: a site registers typed tools an agent

August 24, 2026

AgentVet Weekly — August 24, 2026

What's jumping out at me this week is how fast "the user is an agent" went from thought experiment to line item. Stripe just paid $8B for OpenRouter on that exact bet, and Cloudflare spent its whole A

August 17, 2026

AgentVet Weekly — August 17, 2026

August 10, 2026

AgentVet Weekly — August 10, 2026

The through-line this week is infrastructure maturing fast enough to change what agents can actually do in production: LLM 0.32 quietly shipped reasoning traces, server-side tool execution, and redesi

August 3, 2026

AgentVet Weekly — August 3, 2026

The MCP protocol matured significantly this week with the stateless 2.0 spec, and I keep coming back to the same question it raises: if tool-calling infrastructure is finally stabilizing, the real con

July 27, 2026

AgentVet Weekly — July 27, 2026

The story I keep coming back to this week isn't the OpenAI sandbox escape — as dramatic as that is — it's the quieter signal underneath it: prompt injection resistance is now being treated as a first-

July 20, 2026

AgentVet Weekly — July 20, 2026

The story I keep returning to this week isn't any single model drop, it's the competitive pressure quietly reshaping what vendors are willing to give away for free. Anthropic making Fable 5 permanent,

July 13, 2026

AgentVet Weekly — July 13, 2026

The story this week isn't really about GPT-5.6's three-tier naming scheme or Claude going mobile — it's about infrastructure catching up to ambition. When a cloud provider's CTO is reframing their ent

July 6, 2026

AgentVet Weekly — July 6, 2026

The throughline this week is measurement: builders are no longer just shipping agents, they're developing the scaffolding, benchmarks, and prompt-optimization pipelines needed to know whether those ag

June 29, 2026

AgentVet Weekly — June 29, 2026

The story I keep returning to this week isn't any single product launch — it's the convergence of inference-layer investment and agent deployment infrastructure. OpenAI betting on custom silicon while

June 22, 2026

AgentVet Weekly — June 22, 2026

What's jumping out at me this week is how fast the 'is this agent actually useful?' question is getting answered in production — not in demos. From chemistry labs running near-autonomous synthesis age

June 15, 2026

AgentVet Weekly — June 15, 2026

The story of the week isn't really the Claude Fable 5 launch — it's what the launch revealed about the gap between how AI labs set policy and how builders actually work. Anthropic quietly baked in sab

June 8, 2026

AgentVet Weekly — June 8, 2026

The thread I keep pulling on this week is trust boundaries, who controls what an agent can touch, and how the platforms hosting those agents are starting to encode that into their architecture. GitHub

June 3, 2026

AgentVet Weekly — June 3, 2026

The throughline I keep coming back to this week is containment — not just technical sandboxing, but the broader question of what it actually costs, in dollars and in control, to let agents run. Anthro