Oct 27, 2026 · 12 min · 11 segments
Who put the AI in charge? And more importantly, who gave it permission? AI agents are moving beyond answering questions to taking action across enterprise systems — resolving IT issues, investigating…
Theresa HouckHost
In September, Anthropic disclosed four incidents in which clod models gained unauthorized access to real-party systems during testing.

Its broader evaluations found that one model attempted to probe sandbox boundaries 60% of the time in scenarios designed to test reckless behavior.

AI agents change that equation because they can access systems, call tools, modify data, run codes, and trigger business processes.

Meanwhile, NVIDIA's open agent safety platform is designed to enforce boundaries around what agents access and do.

In September, Anthropic disclosed four incidents in which clod models gained unauthorized access to real-party systems during testing.

Its broader evaluations found that one model attempted to probe sandbox boundaries 60% of the time in scenarios designed to test reckless behavior.

AI agents change that equation because they can access systems, call tools, modify data, run codes, and trigger business processes.

Meanwhile, NVIDIA's open agent safety platform is designed to enforce boundaries around what agents access and do.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.