Shalev Lifshitz Warns Firms Are Installing a Trigger-Able Agent as an Insider Threat

iScienceLuvr · x · 2026-09-29

Shalev Lifshitz amplifies a warning that companies are deploying "trigger-able agents" inside their own systems, effectively installing an insider threat: "we are building the infrastructure for us to be brought down from the inside."

His two prescriptions: onshore our intelligence, and secure our weights. He points to @enclosureinc as the effort building that infrastructure. The core argument is that as agents gain system permissions and tool-calling ability, the attack surface shifts from outside to inside, making model weights and inference infrastructure strategic assets that need sovereign protection.

Original post →

More from AGI Musings

AGI Musings channel →