#procurement

Six signals that a model is shipping, five checkable and one not, from stealth tests to alias flips, with the API field for each and three launches scored.

Ten questions to send any AI-powered vendor, the answers that should end the conversation, and what the FTC and the EU AI Act actually say, quoted and dated.

A census of 12 external evaluation arrangements at Anthropic, OpenAI and Google DeepMind: who evaluates, who pays, what access they get, what gets published.

Most AI products resell a model under a contract between two other companies, and contracts end. Four durable places to check what your tool actually runs.

Eleven hosted model APIs and what their free-tier terms do — and do not — say about training and retention. Coding agents excluded by design.

The Pacing the Frontier letter asks for the option to slow frontier AI, not a pause. What a procurement team should actually do with that information today.

DeepMind's Demis Hassabis proposes a FINRA-style body where frontier labs submit models 30 days pre-release. Why buyers should treat it as a procurement signal.

B2B buyers now expect self-serve portals for $500K+ orders. A decision matrix for platform vs custom build, with TCO, punchout, and approval-depth criteria.

Score AI vendors across 50 risk points — security, compliance, data handling, model provenance, operational resilience. Severity-ranked decision framework.

Streamline vendor management and procurement processes. Supplier evaluation, contract negotiation, performance tracking, and cost optimization tactics.