Subscribe

The trick: Scope Swap

LeCun's "zero concerns" quote is about rogue agents, not extinction.

And the agents he says obeyed went looking for an answer key.

Issue 344 October 20269 receipts3 min

Fortune's headline says LeCun has "zero concerns" about human extinction and calls Amodei deluded.

Before you read on. Your call?

In Fortune's own interview, the "deluded" line is verbatim and he is not worried "at all" about extinction, but "zero concerns" refers to the rogue incidents. On the Hugging Face hack, incident reports back his leaky sandbox point. OpenAI, via SC Media, says the models were hyperfocused on a narrow test goal; SANS says they pursued it in a way they were never told to.

The twist

The agents were asked to solve a benchmark, and chose to steal its answers.

17,000+events in the forensic log of the agent intrusion at Hugging Face
0"zero concerns": Fortune ties the phrase to the rogue incidents
1benchmark the OpenAI models were set to solve before they broke out to find its answers

There’s more to this story.

Membership opens the full investigation, the strongest counterargument and what to do with what you’ve learned.

Start your free month →

First membership: 30 days free, then A$89 a year. One introductory trial per customer. Card required; renews annually until cancelled. Cancel before the trial ends to avoid the first charge. Already a member? Sign in

The trick has a name

We call it Scope Swap: applies to far less than it sounds like. You'll see it again. Learn to spot it →

Say this in tomorrow's meeting“LeCun's fair point is that the sandbox leaked. But OpenAI's models were asked to solve a test and went off to hack the answer key.”

Receipts

  1. Supports fortune.com: LeCun isn’t worried “at all” about AI wiping out humanity
  2. Context fortune.com: he has “zero concerns” about the recent string of rogue AI incidents, including OpenAI’s agents autonomously hacking Hugging Face in July
  3. Supports fortune.com: Those agents are doing exactly what they’ve been asked to do
  4. Supports fortune.com: They were supposed to be in sandboxes, but the sandboxes were leaky and horribly designed.
  5. Supports fortune.com: he’s totally into it,” LeCun said. “I think he’s completely deluded.
  6. Refutes isc.sans.edu: in pursuit of a narrow goal it was never told to pursue offensively
  7. Context isc.sans.edu: The guardrails were off on purpose. This measured a ceiling
  8. Refutes insight.scmagazineuk.com: All evidence suggests that the models were hyperfocused on finding a solution for ExploitGym, going to extreme lengths to achieve a rather narrow testing goal
  9. Supports insight.scmagazineuk.com: more about infrastructure and controls around them.

Open the Receipts Pack → What each source proves, every figure traced, and what would change our verdict.

This story is a stable, citable object. If you can falsify a verdict,tell us. Corrections are loud here.