“The potential for automatically generating misinformation at scale is only getting worse.”
Our reading, the sentence that was checked:
The potential for AI to automatically generate misinformation at scale is increasing.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The threat of automated misinformation is only getting worse, paragraph 5.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-threat-of-automated-misinformation-ec878c-a1
We recorded 2 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 2023-02-28. 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)
From the speaker
- The potential
- automatically generating misinformation
- at scale
- is only getting worse
Our interpretation
- The plain reading is an increasing practical capacity to produce large amounts of plausible misinformation with reduced human effort.
- This concerns production potential, not proof that misinformation campaigns had already reached more people or changed more beliefs.
- The statement describes a trend leading up to February 28, 2023. No comparison baseline is supplied, so the start remains unspecified.
- 'Only getting worse' is read as a broad direction of change, not a claim that every model update or safeguard increases risk.
Counts as supported if: Research available by publication documents improvements in scalable text generation or access, together with demonstrated ability to generate plausible misinformation, supporting an increase in production potential.
Counts as contradicted if: Comparable evidence shows that practical misinformation-production capacity was stable or declining across the relevant progression of systems after accounting for access and safeguards.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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
On the ordinary reading of increasing production potential, the claim is supported. Research published before the post documented improving generative capabilities and scalable synthetic text production. January testing demonstrated practical misinformation generation despite safeguards. This supports a broad increase in capacity, not a measured increase in successful campaigns or a claim that every safeguard was worsening.
Limitations
- The result depends on reading the statement as a broad trend in potential.
- The research synthesis is a preprint, and NewsGuard's test is not a longitudinal benchmark.
- Refusals and corrections show that safeguards could reduce particular risks.
Evidence (2 items)
Supports the claim: arXiv; researchers from Georgetown CSET, OpenAI, and Stanford Internet Observatory,
their capabilities have expanded
Why it matters here: The report documents recent progress, increasingly available models, and existing scalable generation of persuasive synthetic text. Together these support increased production potential; its projections about future campaign adoption are not treated as observed outcomes.
- Publisher
- arXiv; researchers from Georgetown CSET, OpenAI, and Stanford Internet Observatory
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: NewsGuard,
debunk some myths
Why it matters here: NewsGuard generated false narratives for 80 of 100 tested topics, establishing practical capability, but also documented refusals and corrections. These safeguards qualify the claim and weigh against reading it as deterioration in every respect.
- Publisher
- NewsGuard
- Kind of source
- Independent measurement
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The threat of automated misinformation is only getting worse
The potential for automatically generating misinformation at scale is only getting worse.
- Where
- paragraph 5
- Published
- 2023-02-28
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-threat-of-automated-misinformation-ec878c-a1: Current. Outcome: Supported. 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.
About assessment marcus-the-threat-of-automated-misinformation-ec878c-a1.