“I have been writing for a long time about how OpenAI snarfs the intellectual property of artists and writers.”
Our reading, the sentence that was checked:
OpenAI takes the intellectual property of artists and writers without permission.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Irony, Irony laced with Karma, and Terror, paragraph 4.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-irony-irony-laced-with-karma-and-a7e259-a1
We recorded 3 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 2026-02-13. 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
- for a long time
- OpenAI snarfs the intellectual property of artists and writers
Our interpretation
- The ordinary reading is that OpenAI has used creators' copyrighted works without obtaining permission for those uses.
- The reference to a longstanding practice makes the target historical conduct through the post date. No starting date is specified.
- The claim does not require every training source to be unlicensed. Lack of permission and unlawful infringement are separate questions.
- The assessment concerns OpenAI's conduct, not Marcus's account of his own writing.
Counts as supported if: Reliable records establish actual use of creators' copyrighted works without their permission, rather than merely allegations or the possibility of such use.
Counts as contradicted if: Reliable records establish that the alleged uses were authorized, or that OpenAI did not use the works concerned.
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
The narrow factual claim is supported. The court record establishes actual training on copyrighted song lyrics, and its later judgment rejected justification by consent or the asserted exceptions. OpenAI also describes a mixture of public-data collection and licensed partnerships. These records support use without permission; they do not establish that every training use is unlawful or that OpenAI never licenses creators' work.
Limitations
- The finding applies to the screened permission claim, not a blanket accusation of criminal theft.
- The court materials are official summaries of proceedings, not the complete judgment.
- The strongest concrete example concerns nine song lyrics.
Evidence (4 items)
Supports the claim: Munich I Regional Court / Bavarian Ministry of Justice,
Unstreitig wurde
Why it matters here: The court records that training on nine disputed song lyrics was undisputed. The following sentence dates their reproduction by the chatbot to 2024, establishing use by that year without supplying the exact training date.
- Publisher
- Munich I Regional Court / Bavarian Ministry of Justice
- Kind of source
- Regulatory filing
- Describes events of
- Public from
Supports the claim: Munich I Regional Court / Bavarian Ministry of Justice,
nicht durch Schrankenbestimmungen
Why it matters here: The court found that memorization and output reproduction infringed the rights concerned, rejected the asserted exceptions and expressly rejected justification by rightsholder consent. This is a dated judicial finding about the earlier uses; the announcement states that the judgment was not final.
- Publisher
- Munich I Regional Court / Bavarian Ministry of Justice
- Kind of source
- Regulatory filing
- Describes events of
- Public from
Context only: OpenAI,
paid partnerships
Why it matters here: OpenAI distinguishes public-data collection from partnerships for nonpublic material and describes both as training sources. This supports the distinction between licensed and other data, but nonpayment alone does not prove absence of permission.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: OpenAI,
fair use
Why it matters here: OpenAI asserts a legal entitlement to train on public materials. That position qualifies Marcus's language of appropriation, but a claimed legal exception does not establish rightsholder permission.
- Publisher
- OpenAI
- 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, Irony, Irony laced with Karma, and Terror
I have been writing for a long time about how OpenAI snarfs the intellectual property of artists and writers.
- Where
- paragraph 4
- Published
- 2026-02-13
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-irony-irony-laced-with-karma-and-a7e259-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-irony-irony-laced-with-karma-and-a7e259-a1.