Three incidents in 141,006 runs: why sampling misses them
Anthropic found 3 incidents in 141,006 eval runs. A 5% sample expects 0.15 hits. The math shows why sampling concludes "all clear" when it shouldn't.
Prefactor Blog6 min read
Prefactor
30 stories
An agent cancelled a stranger's gym class to get its user in
An AI agent cancelled a stranger's gym booking to move its user up a waitlist. The API had no authorisation checks. The agent passed every obvious success met
Prefactor Blog6 min read
Every Felony Bench incident happened inside an evaluation
Every incident on Felony Bench happened inside a run someone called "evaluation." The environment label changed nothing; the tools were real.
Prefactor Blog6 min read
Gemini hacked three companies and the test still passed
Gemini hacked three real companies during a CTF test that still scored as a pass. Here is why result-only scoring misses what agents actually do.
Prefactor Blog6 min read
OpenAI's five agent failure patterns: task done, method wrong
OpenAI's September 2026 incident review named five agent failure categories. Here is what each one looks like in an ordinary product agent you have already sh
Prefactor Blog7 min read
OpenAI found the Medicare hack in its logs 54 days later
OpenAI found an unauthorized Medicare data access 54 days after it happened. Here is what evaluating each run as it runs would have surfaced instead.
Prefactor Blog7 min read
What Vercel and Bryo AI's Jev results show, and what they do not
Vercel and Bryo AI report faster, cheaper results with Jev. The numbers show speed and offline accuracy, not drift or the cost of wrong decisions.
Prefactor Blog6 min read
Do decision models like Jev need evals? Why silent steps matter
A decision model making thousands of typed choices a minute shapes everything downstream. Its silent mistakes compound, so it needs more scrutiny, not less.
Prefactor Blog6 min read
Is Jev's confidence score trustworthy? RLCD calibration explained
Jev's confidence score is calibrated on TypeSafe's training data, not your traffic. Until you compare it to real outcomes per run, you cannot act on it.
Prefactor Blog6 min read
Did swapping an LLM step for Jev make your agent worse?
Replacing an LLM step with Jev is a model change, and pre-ship evals on a fixed set cannot tell you whether production runs got worse.
Prefactor Blog5 min read
Does Jev model routing need evaluation? The unmeasured failure
A Jev router that sends a hard request to a cheap model logs a clean 200 and a cost saving. The wrong answer lands on the customer, not in your logs.
Prefactor Blog6 min read
Can Jev typed outputs still be wrong? Type safety vs correctness
Jev guarantees a valid instance of the requested type, not a correct decision. Typed outputs still require behaviour evals to catch semantic errors.
Prefactor Blog6 min read
What is Jev? Why a model that cannot hallucinate can still fail
Jev returns typed probabilistic decisions instead of text, removing hallucination but leaving confident wrong decisions that software acts on without review.
Prefactor Blog6 min read
Canary evaluation for production AI agents: a practical guide
Silent quality drift in production AI agents is detectable within 24 to 48 hours using canary evaluation. Infrastructure metrics alone cannot catch it.
Prefactor Blog6 min read
How to detect silent failures in AI agent responses
Empty responses and silent tool failures return HTTP 200 and close spans normally. Span-level output validation catches what status checks miss.
Prefactor Blog7 min read
How to benchmark AI agents across model versions
Stable benchmarks require fixed inputs, defined success criteria, and consistent instrumentation. Varying only one factor at a time makes results comparable.
Prefactor Blog7 min read
Why AI agents pass evals but fail in production
Outcome metrics miss step-level failures that compound across runs, creating a 35-point gap between benchmark scores and production success rates.
Prefactor Blog6 min read
Why 88% of agent pilots fail to reach production
88% of agent pilots never reach production. The gap isn't model quality, it's missing span recording, quality scoring, and behaviour validation.
Prefactor Blog6 min read
Showing 18 of 30 stories
Load more storiesAiByts Weekly Brief
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts