Field notes on breaking AI.
Research, engineering deep-dives and the occasional strong opinion. No fluff, no listicles about "the future of AI".
Your eval suite is lying to you: 7 ways LLM tests pass when they shouldn't
A green dashboard is not a safe model. Here are the seven failure modes we see most often when we audit customer eval suites, and how to fix each one.
Prompt injection is a supply-chain problem now
When your agent reads email, tickets and web pages, every one of those sources becomes part of your attack surface. We need to start treating it that way.
Guardrails at 28ms: how we built a proxy that doesn't slow you down
A guardrail that adds a second of latency gets turned off. Here's the architecture behind Probe Guard's p95 under 30ms, and the trade-offs we made to get there.
The red-team playbook we give every new customer
The exact five-phase process our researchers follow in the first week of an engagement, so you can run a version of it yourself.
Why we'll never train a foundation model
Every AI company is pressured to build its own model. We made the opposite promise on day one. Here's why neutrality is the product.