Morning Edition · Sunday, November 22, 2026Published at 1:11 AM EST · New York
The attack used no prompt injection. Separately, a Google DeepMind paper reports that automated agent attackers can still steer models that look robust on public prompt-injection leaderboards.

NebuSec, a security research firm, published two write-ups on November 20 describing a full compromise of Muse. Muse is the consumer agent Meta launched on September 8, built on the Muse model line that Meta also sells to developers through…
Track frontier labs, chips, export controls, model releases, regulation, and AI infrastructure.
Track all trends and read every AI and Crypto article for $19/month.
Part of a tracked trend
Oversight and Evaluation Lag Accelerating AI Capabilities
Over the next 3-6 months, evidence mounts that governance, evaluation, and agent-safety methods are failing to keep pace with capability growth, driving investment in interpretability, agent-manipulation benchmarks, and institutional-reform proposals.
Start a discussion in Townsquare.
More from this edition
Comments
0No comments yet.