APIR/Blog
APIR
APIR Intelligence · the blog

Field notes

Field notes on showing an AI’s work: the papers an agent carries, the log book behind them, sealing what is real and checking the numbers a buyer is handed.

The SealLatest

Before Your Agent Calls an API, Ask Who Stands Behind It

An agent calls a service on trust. An API seal records a declaration by whoever could make the domain answer, over the exact spec file, on a date. The lookup is free and needs no key.

Sep 28, 2026 4 min readRead article
The Seal

Seal What's Real. Then Check Anything Against It.

A dated declaration over the exact bytes of a file, on a chain an outside authority timestamps every night, checkable with APIR switched off. And the check on the other side: drop any file and see whether anyone sealed it.

Sep 28, 2026 5 min read
The Record

A Trust Page You Can Check

Most trust pages are promises and a badge. Ours holds no badge. Every line carries one of three marks: checkable, our word, or not held.

Sep 28, 2026 3 min read
The Check

Two Tests the Seller's Numbers Must Pass

Before you trust an information memorandum, run the two arithmetic tests every buyer should: the tax return against the management accounts, and the cash cycle. Free in your browser, nothing uploaded.

Sep 28, 2026 5 min read
The Record

Everyone Can Build an Agent Now. Nobody Can Vouch for One.

Standing up an autonomous agent takes ten minutes. Finding out what one is, who is accountable for it, and what it has done before is still a research project. That gap is where the liability sits.

Aug 15, 2026 6 min read
The Record

How Your AI Decided Is Now Discoverable

In March 2026 a federal magistrate ordered discovery into how an automated review tool was built and what it did. How your AI reached a decision is now something a court can compel, and staff recollection is a poor substitute for a decision level record.

Aug 15, 2026 6 min read
Regulation

Nobody Preempted Anything: The US AI Disclosure Map as at August 2026

There are 109 AI laws across 29 states and no federal preemption. A state by state map of who each US AI disclosure rule actually binds, and why every one of them ends up being an evidence problem.

Aug 15, 2026 6 min read
Stop reading, start checking

Every agent in your stack is either papered or it is not.

Check any agent that contacts you, free and with no account. Or hire your first one with papers, free.

Check an agent Hire your first, free

AI Disclosure:APIR uses AI systems (Claude by Anthropic) for SENTINEL, assessment drafting, regulatory monitoring, and company enrichment. AI-generated outputs are labelled as such.

Full disclosure →Report AI Incident

Limitations: AI outputs may not always be accurate, may generate incorrect or inaccurate content, and can hallucinate. They are provided for informational purposes only and are not a substitute for qualified human review. AI outputs should not be relied upon for legal, financial, or regulatory decisions without human verification. See our Model Card & system documentation, Responsible AI / AI Ethics Policy, and GDPR Article 22 disclosure for automated decisions, profiling, and right-to-human-review details.