2026-03-12 23:45 UTC
Overpermissioning was already a problem. Now, imagine those permissions assigned to AI agents operating at machine speed.
At #RSAC2026, Nancy Wang, CTO, @1password, joins Fotis Chantzis, Agent Security Lead at @OpenAI, to unpack ⤵️
🔹 Why legacy access models break in AI-driven environments
🔹 How agent overpermissioning compounds risk
🔹 What leaders must rethink now
📅 March 25 | 3 PM PDT
👉 Register: https://bit.ly/4b4zCQj
Replies (1)
-
@joshg@mathstodon.xyz 2026-03-13 00:01
@1password@1password.social here's a great solution: ffs, don't hand the controls over to these fundamentally flawed systems. they're only being offered to us because they're hoping we'll train them to not suck by letting them make mistakes all over our stuff, not theirs