“Essentially the same argument applies to Claude’s constitution, which amounts to a set of built-in guardrails. OpenAI has those, too!”
Our reading, the sentence that was checked:
Claude's constitution is basically a set of built-in guardrails, and OpenAI's models have guardrails too.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Is the US military actually afraid of Claude? A new theory of why Anthropic was labeled a supply chain risk., paragraph 13.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-is-the-us-military-actually-afraid-f4d323-a1
We recorded 10 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 2026-03-12. 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
- Claude’s constitution
- a set of built-in guardrails
- OpenAI has those, too!
- Just constructed differently
Our interpretation
- Read guardrails in its ordinary broad sense: behavioral guidance and safety limits incorporated into model training or operation.
- Built-in does not require an infallible software filter or identical implementations at the two companies.
- Read amounts to as a functional summary. An exhaustive claim that the constitution contains only restrictions would be materially stronger.
- Use the current calendar quarter through publication, January 1–March 12, 2026, as the reporting window for the ongoing design state.
Counts as supported if: Anthropic documents that its constitution shapes trained behavior and safety limits, and OpenAI documents implemented behavioral safeguards in its models.
Counts as contradicted if: The constitution is merely an unused statement without a role in model behavior, or OpenAI's relevant models lack implemented behavioral 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
Both companies document behavioral constraints incorporated into their models. Anthropic uses the constitution in training, while OpenAI describes trained adherence to safety policies. This supports the ordinary functional comparison. The constitution also addresses values and judgment, so calling it only a collection of restrictions would be incomplete.
Limitations
- The outcome depends on reading guardrails broadly and amounts to as shorthand.
- The sources establish design and training practices, not perfect compliance.
- The evidence does not establish equal safety, equal military suitability, or identical implementation.
Evidence (3 items)
Supports the claim: Anthropic,
its content directly shapes Claude’s behavior.
Why it matters here: Anthropic directly connects the constitution to model training. The same announcement describes safety priorities and hard behavioral constraints, supporting the broad guardrails reading.
- Publisher
- Anthropic
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Supports the claim: OpenAI,
guidelines and model policies we’ve set, helping them act in line with our safety expectations.
Why it matters here: The system card describes training that enables compliance with behavioral policies and safety expectations. This establishes OpenAI's implementation of guardrails in the broad sense used in the contract.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: Anthropic,
identify the best possible action in situations that such rules might fail to anticipate.
Why it matters here: The constitution includes broader judgment and values as well as hard constraints. This qualifies Marcus's shorthand and weighs against treating it as an exhaustive description of the document.
- Publisher
- Anthropic
- 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, Is the US military actually afraid of Claude? A new theory of why Anthropic was labeled a supply chain risk.
Essentially the same argument applies to Claude’s constitution, which amounts to a set of built-in guardrails. OpenAI has those, too!
- Where
- paragraph 13
- Published
- 2026-03-12
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-is-the-us-military-actually-afraid-f4d323-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-is-the-us-military-actually-afraid-f4d323-a1.