Pundit ClaimsAI claims by Gary Marcus and Ed Zitron, with the evidence

“Every big tech company is desperate for training data, and we can expect more companies following the same sly opt-out playbook, hiding what is going in obscure settings and Terms of Service nobody reads.”

PSA: Microsoft may be training on your private data without your knowledge, paragraph 7. The full statement is at the source.

Our reading, the sentence that was checked:

More companies will follow an opt-out playbook to collect training data, hiding it in obscure settings and Terms of Service.

Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of

Where and when Gary Marcus wrote it

Marcus on AI, PSA: Microsoft may be training on your private data without your knowledge, paragraph 7.

Said
Source observed
live page, full text, public
Evidence as of
the cutoff for this assessment
Assessment recorded
assessment marcus-psa-microsoft-may-be-training-on-930bb2-2-a2

We recorded 6 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 2024-11-26. This page does not claim the wording was captured then.

What the claim commits to

Kind
Forecast
Deadline, as the speaker wrote it
None stated. This site does not add one.
Deadline state when assessed
No deadline stated
Deadline state on 2026-10-29
No deadline stated (the release’s reference date)

From the speaker

  • we can expect more companies
  • following the same sly opt-out playbook
  • hiding what is going in obscure settings and Terms of Service nobody reads

Our interpretation

  • The graded forecast concerns additional companies adopting the practice after November 26, 2024.
  • More companies requires multiple additional companies; two is a minimum, not a claim about widespread prevalence.
  • Hiding is assessed through disclosure and interface design, not an inferred motive. Default training alone is insufficient.
  • Nobody reads is rhetorical emphasis on poor visibility, not a literal claim that no person reads terms.
  • The prefatory assertion about every company's desperation is not part of this extracted forecast.

Counts as supported if: At least two additional companies introduce default inclusion of user data in AI training after the post, with evidence that notice and opt-outs were confined to obscure settings or terms rather than clearly presented to affected users.

Counts as contradicted if: A decisive development rules out the predicted additional adoption. Failure to find examples by an arbitrary date cannot contradict an undated forecast.

Retrospective: these criteria were written after the outcome was public.

Current assessment

Did the forecast come true?

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 deadline stated
Assessment recorded
Policy version
2026-10-28.7

Rationale

Reuters reported on January 30, 2026 that Starlink's January 15 policy update permitted AI training unless users opted out, but the report does not establish what notices affected users received [e1]. Vercel's March 17, 2026 announcement described an opt-out deadline of March 31 [e2]. A March 8, 2025 account described Pinterest's planned default inclusion under a policy effective April 30, with a settings opt-out and advance notice [e3]. These records do not establish two companies introducing the complete practice of default training inclusion with obscure disclosure, or a decisive development ruling out further adoption [e1, e2, e3].

Limitations

  • Original Starlink user notices and interface records are missing.
  • A public announcement alone does not establish the visibility of notice to every affected user.
  • Pinterest's advance notice is described in a secondary account.
  • Policy permission does not prove that training occurred.
  • The forecast has no deadline, and the reviewed examples do not establish what happened across all companies.

Evidence (3 items)

  • Context only: Reuters, syndicated by Yahoo Finance,

    what data

    Why it matters here: Reuters documents a new opt-out training permission and unclear data scope. It does not establish whether affected users received prominent notices or whether training had actually begun.

    Publisher
    Reuters, syndicated by Yahoo Finance
    Kind of source
    News report
    Describes events of
    Public from
  • Context only: Vercel,

    opt out

    Why it matters here: The announcement specifies default inclusion for Hobby and trial accounts, with paid Pro and Enterprise excluded by default. Its advance announcement prevents treating the policy alone as proof of concealed disclosure.

    Publisher
    Vercel
    Kind of source
    Company disclosure: the company describing itself, not an independent finding
    Describes events of
    Public from
  • Context only: Lindsey Gamble,

    notify users in advance

    Why it matters here: The account describes default inclusion and a settings opt-out, but also advance notice. This makes Pinterest an uncertain example of the complete concealed-disclosure forecast.

    Publisher
    Lindsey Gamble
    Kind of source
    Secondary summary
    Describes events of
    Public from

What would settle it

Dated records establishing that at least two additional companies introduced default AI-training inclusion after November 26, 2024, with notices and opt-outs confined to obscure settings or terms. Each case needs the introduction date, data scope, default, opt-out path, and user notices. A decisive development ruling out further qualifying adoption could instead contradict the forecast.

Where it was said (1)

Said once in the sources this register covers.

  • : Marcus on AI, PSA: Microsoft may be training on your private data without your knowledge

    Every big tech company is desperate for training data, and we can expect more companies following the same sly opt-out playbook, hiding what is going in obscure settings and Terms of Service nobody reads.

    Published
    2024-11-26
    Source observed
    , live page
    Access
    full text, public

History

Assessment versions

Recorded events

  1. Assessment publishedBy this site

    First assessment published after both reviews and the final-text check passed.

    About assessment marcus-psa-microsoft-may-be-training-on-930bb2-2-a2.