Monday 16 November
Global

Your region, kept on this device.

Microsoft Research shows AI model neurons can be edited to hide sleeper agent backdoors

Severity
Medium
Sources
1

demonstrates weight-space backdoors in which fewer than 10 neurons are edited to inject a short payload that fires on a trigger string, while the model passes standard evaluations; the attack needs only inference-level compute. with Opus 4.5 discovered most of the threat autonomously. Defenders should diff checkpoints against their declared base model, flag statistical outlier tensors, source weights only from original publishers with verified hashes, run agents with least privilege in isolated sandboxes, restrict egress, and require human approval for high-impact tool calls. No CVE is referenced.

For security analysts · Written by AI from 1 source · Wrong?

What to do

  1. Diff every model checkpoint against its declared base model before deployment and after each update.
  2. Source checkpoints only from the original publisher and verify the published file hashes.
  3. Run agents with least privilege and require human approval for high-impact tool calls.
  4. Restrict agent network egress so a triggered model cannot open arbitrary outbound connections.

Written by AI from the sources, checked against them

Full story timeline · 1 update

  1. Tue 10 NovemberDetecting sleeper agents hiding in AI model neuronsMicrosoft Security Blog

Who reported · 1 outlet

Microsoft Security Blog

Who and what is in it

Products

Report a problem