“As a society, we still don’t really have any plan whatsoever about what we might do to mitigate long-term risks.”
Our reading:
Society has no real plan to mitigate long-term AI risks.
Assessment: Unresolved
What happened
Was it true for the period described?
Outcome: Unresolved
We checked, and the evidence so far doesn’t settle it.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
Concrete public preparation existed: the UK was building government frontier-risk assessment capacity, and the White House had coordinated safety commitments. Those records weaken a literal claim that nothing was being planned. But Marcus's context concerns collective machinery for addressing serious future threats, and neither record establishes that broader response plan. The international summit agenda also shows practical arrangements still being developed. These examples cannot prove the absence of every qualifying public plan. Calling the claim contradicted would require treating partial preparations as society's plan; calling it supported would require stronger evidence of absence.
Caveats
- The geographic scope and minimum content of a real plan are unspecified.
- The sources document selected public initiatives, not a comprehensive inventory.
- The schema has no government-policy-disclosure source type; company_disclosure is used for official government announcements and reports.
Evidence (3 items)
Context only: UK Department for Science, Innovation and Technology,
cybersecurity and frontier AI
Why it matters: The government report describes a public frontier-risk team, national-security participation, and technical partnerships. This is material evidence of collective preparation, but the report does not establish what public authorities would do in response to a dangerous finding.
- Publisher
- UK Department for Science, Innovation and Technology
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Context only: The White House,
internal and external security testing
Why it matters: The fact sheet describes government-coordinated commitments covering biological and cybersecurity risks, while an executive order and legislation remained prospective. These precautions count against reading the claim as no preparation at all, but do not independently establish a public response plan for advanced-AI threats.
- Publisher
- The White House
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Context only: UK Department for Science, Innovation and Technology,
agreeing practical next steps
Why it matters: The announcement places practical international next steps and a longer-term roadmap on a future summit's agenda. That supports the concern that arrangements were still developing, but it also acknowledges existing initiatives and cannot establish that no qualifying public plan existed.
- Publisher
- UK Department for Science, Innovation and Technology
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Probably right?
One AI thinks this is probably right, and the other says it’s too early to tell. The evidence isn’t conclusive, so the verdict above stays unresolved. It counts only when you pick Speculative under “How sure?”.
- Anthropic’s AI: The Sept 2023 White House fact sheet calls voluntary commitments a bridge, with an executive order still being drafted. The UK taskforce was doing research, and its summit was to agree next steps later. So contemporary arrangements were preparatory, not an adopted plan.
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Where we had to interpret
- The plain reading is that society lacked a concrete collective or public plan for managing serious risks from future advanced AI. A company's individual safety policy does not establish such a plan.
- The surrounding discussion distinguishes proposals and voluntary assurances from public arrangements capable of addressing threats. A qualifying plan should connect risk assessment to specified preventive or response measures and identify who would carry them out. It need not eliminate every risk or consist entirely of binding law.
- The context emphasizes the United States but also discusses international institutions. Neither an exclusively American scope nor a requirement for one worldwide plan is explicit.
- This is an observation about arrangements existing when the post appeared. With no regular reporting period for public policy, the target period is September 2023 through the September 21 publication date. Later developments cannot disprove the historical claim.
Right if: Reliable contemporary accounts establish that relevant public arrangements remained proposals, voluntary assurances, or preparatory research, without an adopted collective plan connecting assessment of serious future AI threats to concrete preventive or response action.
Wrong if: Contemporary records establish an adopted public or collectively governed plan, applicable at the societal level under discussion, with responsible actors and concrete arrangements linking detection of serious future AI threats to prevention or response. One company's policy, a legislative proposal, or a research agenda alone is insufficient.
We wrote this test after the outcome was known.
What would settle it
A contemporary public strategy with adopted responsibilities and concrete prevention or response arrangements for advanced-AI threats, or a sufficiently comprehensive contemporary review establishing that relevant public authorities had no such arrangements.
We found 10 claims in this post. See them all.
Full record
Source and dates
Marcus on AI, Six months after the Pause letter, AI is still erratic — and still something we should still be deeply concerned about, paragraph 13.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-29. The post may have been edited after 2023-09-21.
Where it was said (1)
Said once in the posts we read.
: Marcus on AI, Six months after the Pause letter, AI is still erratic — and still something we should still be deeply concerned about
As a society, we still don’t really have any plan whatsoever about what we might do to mitigate long-term risks.
- Where
- paragraph 13
- Published
- 2023-09-21
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Unresolved. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
A reviewer said our test wasn’t fair to what was actually said.