“OpenAI had privileged access to data their competitors didn’t have, but didn’t acknowledge this.”
Our reading, the sentence that was checked:
OpenAI had privileged access to FrontierMath data that competitors lacked and did not acknowledge it.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, OpenAI’s dirty December o3 demo doesn’t readily replicate, paragraph 2.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-openais-dirty-december-o3-demo-doesnt-fc328e-a1
We recorded 8 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-04-23. 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
- OpenAI had privileged access to data
- their competitors didn’t have
- didn’t acknowledge this
- FrontierMath
Our interpretation
- The claim concerns access and disclosure when the December 20, 2024 FrontierMath result was presented.
- Privileged access means access to benchmark problems or solutions beyond the access available to competing developers.
- The claim alleges undisclosed access, not that OpenAI trained on the answers or fabricated its score.
- Later disclosure does not negate nondisclosure at the announcement.
Counts as supported if: Primary records establish preferential access to FrontierMath data and that the access arrangement was not disclosed with the December announcement.
Counts as contradicted if: OpenAI lacked preferential access at that time, competitors had equivalent access, or the announcement clearly disclosed the arrangement.
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
Epoch's primary clarification establishes preferential access and says the data-access arrangement was not clarified when the partnership was announced around the o3 launch. This supports Marcus's access-and-disclosure claim. It does not establish that OpenAI used the data for training or that the benchmark score was invalid.
Limitations
- Epoch could evaluate competitors' models without giving their developers the underlying dataset.
- The arrangement included holdout protections.
- Sponsorship disclosure and data-access disclosure were separate; this finding concerns access.
Evidence (1 item)
Supports the claim: Epoch AI,
we didn’t clarify the data access
Why it matters here: Epoch retrospectively states that OpenAI owned the commissioned problems and could access problems and solutions subject to a holdout exception, while sharing with other parties required OpenAI's written permission. It distinguishes the partnership disclosure around the o3 announcement from the data-access arrangement, which it acknowledges was not clarified then.
- Publisher
- Epoch AI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, OpenAI’s dirty December o3 demo doesn’t readily replicate
OpenAI had privileged access to data their competitors didn’t have, but didn’t acknowledge this.
- Where
- paragraph 2
- Published
- 2025-04-23
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-openais-dirty-december-o3-demo-doesnt-fc328e-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-openais-dirty-december-o3-demo-doesnt-fc328e-a1.