“I warned that LLMs would be used by bad actors, and they increasingly are.”
Our reading, the sentence that was checked:
LLMs are increasingly being used by bad actors.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, TED, AI, and the test of time, paragraph 6.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-ted-ai-and-the-test-of-time-e30d9b-a1
We recorded 11 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 2024-04-15. 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
- LLMs would be used by bad actors
- they increasingly are
Our interpretation
- This assesses the claimed increase in actual malicious use, not the earlier forecast.
- The comparison is with the previous TED year. Use April 2023 through April 15, 2024; the exact baseline date is not supplied.
- Malicious use includes research, coding and deceptive-content assistance for harmful operations. It need not produce a successful attack.
- An increase in published reports, interest or future risk alone does not establish an increase in actual use.
Counts as supported if: Comparable observations or a substantiated longitudinal threat assessment establish growing actual LLM use by malicious actors during the target period.
Counts as contradicted if: Comparable observations establish that actual malicious LLM use was flat or declining over the relevant comparison period.
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
Actual malicious use is documented. The stronger word 'increasingly' remains unsettled: the provider investigation gives examples without a comparable baseline, and CrowdStrike describes sparse observations and incomplete visibility. TIME's report of increasing experimentation favors Marcus's claim, but does not close that measurement gap. These records do not establish that use was flat or declining either.
Limitations
- Malicious use can remain undetected.
- An exact numerical growth rate is unnecessary, but a credible comparison is still needed.
- Later increases would not by themselves establish the trend before this post.
Evidence (3 items)
Context only: OpenAI,
coding errors
Why it matters here: The investigation establishes malicious use detected by February 2024, including five state-affiliated groups. It provides no comparable earlier count establishing an increase; individual usage dates are unspecified.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: CrowdStrike,
rarely observed
Why it matters here: CrowdStrike reports sparse observations and incomplete visibility. This limits confidence in a trend inference, but neither proves declining use nor rules out growth from a small base.
- Publisher
- CrowdStrike
- Kind of source
- Independent measurement
- Describes events of
- Public from
Context only: TIME,
going to get worse
Why it matters here: TIME describes increasing experimentation, which favors Marcus's direction of travel. Its quoted warning concerns future growth, while the inspected primary report does not supply a comparable historical measure of actual LLM use.
- Where
- Paragraph 3
- Publisher
- TIME
- Kind of source
- News report
- Describes events of
- Public from
What would settle it
Comparable dated observations, or a substantiated retrospective threat assessment, showing whether actual malicious LLM use grew between the previous TED year and April 15, 2024.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, TED, AI, and the test of time
I warned that LLMs would be used by bad actors, and they increasingly are.
- Where
- paragraph 6
- Published
- 2024-04-15
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-ted-ai-and-the-test-of-time-e30d9b-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-ted-ai-and-the-test-of-time-e30d9b-a1.