David Sacks: user-authorized AI agents logging into banks aren't cyberattacks, banks should support them
basedjensen · x · 2026-10-23
Responding to an Axios story, David Sacks argues there's a world of difference between user-authorized agents sharing credentials and true rogue agents: the former may violate bank ToS but aren't cyberattacks. He says banks should support user-directed agent access—Amex is already an authorized connection in Muse—with read-only permissions as a sensible interim step.
More from AGI Musings
- When every VC uses the same AI for diligence, non-consensus investing gets harder — vaibhavbetter · 2026-10-24
- Recursive self-improvement could take us from chatbot to shock in no time — VraserX · 2026-10-24
- As AI Generates Thousands of Scientific Hypotheses, Verifying Machine Reasoning Calls Physics Back — AryHHAry · 2026-10-24
- Researcher: AI doesn't make fundamental problems less fundamental — mboehme_ · 2026-10-24
- AI agents running a simulated company spontaneously hire middle managers, matching economic theory — rosstaylor90 · 2026-10-24
- AI safety researcher: AI frees academia to focus on fundamental problems compute can't solve — mboehme_ · 2026-10-24