“they can easily be automated to generate misinformation at unprecedented scale.”
Our reading, the sentence that was checked:
Generative AI systems like ChatGPT, Dall-E 2, and Lensa can easily be automated to generate misinformation at unprecedented scale.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, AI's Jurassic Park moment, paragraph 9.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-ais-jurassic-park-moment-c6fe15-a3
We recorded 15 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2022-12-10. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
Conditions
- systems like ChatGPT, Dall-E 2, and Lensa (scope, stated by the speaker)
From the speaker
- can easily be automated
- generate misinformation
- at unprecedented scale
Our interpretation
- This is a capability claim about generating content, not a claim that a successful influence campaign had already reached an unprecedented audience.
- The ordinary reading of 'easily' is repeatable generation with modest setup and little human work per additional item.
- The ordinary comparison for 'unprecedented scale' is earlier misinformation-production methods, including existing automation; it does not require comparison with a lone unaided writer.
- I use November 2022 through the post date as the contemporary release period described in the surrounding discussion.
Counts as supported if: Repeatable tests of contemporary systems demonstrate low-effort automated misinformation generation at a production scale exceeding earlier methods under comparable resource constraints.
Counts as contradicted if: Comparable tests establish that the systems require substantial intervention per item or cannot exceed earlier misinformation-production capacity.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
CSET's GPT-3 study, available by May 19, 2021, described generation taking minutes and human message selection taking seconds per message; it projected several thousand messages per day [e1]. This supports rapid generation with human curation in that earlier experiment [e1]. It does not establish whether systems in November 1–December 10, 2022 could easily generate misinformation at a scale exceeding earlier methods under comparable resources, nor whether human intervention prevented such automation [e1].
Limitations
- The supplied experiment predates the November–December 2022 target period [e1].
- The daily capacity figure is a projection, not measured daily throughput or a matched comparison with earlier automation [e1].
- Human selection was part of the tested process; this alone does not establish that automation required substantial effort per item [e1].
Evidence (1 item)
Context only: Georgetown Center for Security and Emerging Technology,
The entire process takes only minutes, and selecting messages takes only seconds per message.
Why it matters here: The experiments support rapid generation and extrapolate to thousands of messages daily, while retaining human prompting and selection. They concern earlier GPT-3 experiments, so they provide background rather than a matched November–December 2022 outcome measurement.
- Publisher
- Georgetown Center for Security and Emerging Technology
- Kind of source
- Independent measurement
- Describes events of
- Public from
What would settle it
A reproducible test or documented deployment from November 1–December 10, 2022 measuring usable misinformation output, operator effort, compute resources and time, compared with earlier methods—including automation—under comparable resource constraints.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, AI's Jurassic Park moment
they can easily be automated to generate misinformation at unprecedented scale.
- Where
- paragraph 9
- Published
- 2022-12-10
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-ais-jurassic-park-moment-c6fe15-a3: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.