“It’s a pattern.”
Our reading, the sentence that was checked:
Sam Altman's lack of candor, seen in his statements to the Senate about OpenAI equity, his lies about Helen Toner and his lies about Scarlett Johansson, is a pattern.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The OpenAI board was right, paragraph 28.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-openai-board-was-right-147e9e-a2
We recorded 8 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2024-05-21. 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
- referring to the preceding examples of Altman not being consistently candid (scope, stated by the speaker)
From the speaker
- It’s a pattern.
- his fudges to the Senate
- his lies about Helen Toner
- his embarrassing lies about Scarlett Johansson
Our interpretation
- The ordinary reading is repeated material deception across the preceding examples, not simply repeated controversy.
- A pattern requires at least two distinct, sufficiently established instances. It need not establish that every cited example is correct.
- The reference to earlier examples supplies scope, not a conditional forecast.
- The historical period covers the Senate testimony through the post. Later conduct cannot independently establish the pattern asserted at publication.
- Personal candor is the assessed subject; metric_scope is null.
Counts as supported if: At least two of the named episodes are established as material deception through records, admissions or sufficiently corroborated accounts.
Counts as contradicted if: Reliable records resolve the named episodes such that fewer than two involved material deception. Failure to prove an episode is not itself proof that no pattern existed.
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
On May 16, 2023, Altman told the Senate, “I have no equity in OpenAI”; a December 19, 2024 report describes indirect fund exposure, but these records do not establish that the answer constituted material deception [e1, e2]. In her May 28, 2024 interview, Toner alleged false statements concerning her in October 2023; the item does not supply the underlying exchanges or independent corroboration sufficient to establish their falsity [e3]. Reporting on May 23, 2024 describes casting records and the Sky actress’s account that she used her natural voice without instructions to imitate Johansson, weakening that allegation without conclusively resolving it [e4]. The packet neither establishes two distinct episodes of material deception during May 16, 2023–May 21, 2024 nor reliably establishes that fewer than two occurred [e1–e4].
Limitations
- The equity evidence does not establish whether the Senate answer was a misleading omission or material deception.
- Toner’s account is participant testimony; the packet lacks complete disputed communications and sufficiently independent corroboration.
- The casting evidence does not conclusively establish whether Altman’s denial was false.
- The later equity report provides context for the Senate answer, not a separate deceptive act within the assessed period.
Evidence (4 items)
Supports the claim: Tech Policy Press,
no equity
Why it matters here: The unqualified answer is consistent with Marcus's criticism of an omission. Establishing a broader pattern still requires another sufficiently established deceptive episode.
- Publisher
- Tech Policy Press
- Kind of source
- Secondary summary
- Describes events of
- Public from
Supports the claim: TED,
lying to the board
Why it matters here: Toner alleges repeated misrepresentation and specifically dates efforts to remove her through false statements to late October 2023. This is direct participant testimony, but it does not publish the underlying exchanges or independently establish their falsity.
- Publisher
- TED
- Kind of source
- News report
- Describes events of
- Public from
Context only: TechCrunch,
indirect investment
Why it matters here: The report distinguishes indirect exposure from direct ownership and says the YC investment preceded full-time employment. Its later disclosure is context for interpreting the Senate statement, not a new in-period deceptive act.
- Publisher
- TechCrunch
- Kind of source
- News report
- Describes events of
- Public from
Context only: Ars Technica,
voice
Why it matters here: Reported casting records and the agent's account weaken the Johansson example. They do not disprove a pattern arising from other episodes, and do not conclusively resolve private selection motives.
- Publisher
- Ars Technica
- Kind of source
- Secondary summary
- Describes events of
- Public from
What would settle it
Authenticated records or independently corroborated accounts resolving the named episodes: the equity interests and disclosures relevant to the Senate answer, the disputed communications concerning Toner, and the casting and selection communications concerning Johansson. These would need to establish at least two instances of material deception, or reliably resolve enough episodes to establish that fewer than two occurred.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The OpenAI board was right
It’s a pattern.
- Where
- paragraph 28
- Published
- 2024-05-21
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-openai-board-was-right-147e9e-a2: 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.
About assessment marcus-the-openai-board-was-right-147e9e-a2.