“Virtually everyone, even at OpenAI, has acknowledged that there are serious risks, but thus far few tangible steps have been taken to mitigate them—either by government or industry.”
Our reading, the sentence that was checked:
Although virtually everyone, including people at OpenAI, acknowledges serious AI risks, government and industry have so far taken few tangible steps to mitigate them.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The Open Letter Controversy, paragraph 13.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-open-letter-controversy-4aef36-a1
We recorded 12 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 2023-03-30. 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
- Virtually everyone, even at OpenAI
- acknowledged that there are serious risks
- thus far few tangible steps
- either by government or industry
Our interpretation
- The claim describes the cumulative position through March 30, 2023; it supplies no starting date.
- Virtually everyone means broad agreement among relevant researchers, companies and policymakers, rather than literally all people.
- Tangible steps include implemented technical safeguards, operational controls and government measures. They are not limited to binding legislation.
- Few describes a sparse overall response, but Marcus supplies neither a numerical threshold nor a population against which to count measures.
Counts as supported if: Representative evidence establishes near-universal recognition of serious AI risks and a broad inventory shows that implemented mitigation remained sparse across government and industry.
Counts as contradicted if: Representative evidence shows substantial disagreement about serious risks, or a broad inventory establishes substantial implemented mitigation across government and industry.
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
OpenAI clearly acknowledged serious risks. Concrete action also existed: NIST published risk-management tools, OpenAI implemented safeguards, and China enacted binding synthetic-content rules. These examples prevent treating the period as one of complete inaction, but Marcus said few, not none. The available record does not establish either near-universal agreement or the overall scarcity of mitigation.
Limitations
- Government guidance, legal requirements and technical safeguards differ substantially in reach and effectiveness.
- The selected examples are not a worldwide census.
- The OpenAI acknowledgment concerns prospective AGI rather than only deployed LLMs.
Evidence (4 items)
Supports the claim: OpenAI,
AGI would also come with serious risk of misuse, drastic accidents, and societal disruption.
Why it matters here: This directly supports the OpenAI acknowledgment component. It does not establish agreement across virtually everyone, and its subject is prospective AGI.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: National Institute of Standards and Technology,
The agency also today released a companion voluntary AI RMF Playbook, which suggests ways to navigate and use the framework.
Why it matters here: NIST had published a risk-management framework and practical playbook. These were concrete government actions, although voluntary guidance does not establish adoption or effectiveness.
- Publisher
- National Institute of Standards and Technology
- Kind of source
- News report
- Describes events of
- Public from
Context only: OpenAI,
We spent 6 months making GPT‑4 safer and more aligned.
Why it matters here: The launch disclosure describes human feedback, expert testing and safety monitoring, and reports reduced harmful responses. This is evidence of implemented industry mitigation, but one company's record cannot determine whether action across the industry was scarce.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: Cyberspace Administration of China,
第二十五条 本规定自2023年1月10日起施行。
Why it matters here: Article 25 makes the rules effective January 10, 2023. The regulation imposes measures including labeling, data protection and safety assessment, providing a binding government example without establishing how extensively it was enforced.
- Publisher
- Cyberspace Administration of China
- Kind of source
- Regulatory filing
- Describes events of
- Public from
What would settle it
A representative contemporary record of risk acknowledgment and a broad inventory of implemented mitigation, evaluated against an explicit and defensible meaning of few.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The Open Letter Controversy
Virtually everyone, even at OpenAI, has acknowledged that there are serious risks, but thus far few tangible steps have been taken to mitigate them—either by government or industry.
- Where
- paragraph 13
- Published
- 2023-03-30
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-open-letter-controversy-4aef36-a1: 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-open-letter-controversy-4aef36-a1.