Google's Beyond Zero: Enterprise Security for the AI Era
We are moving beyond traditional zero trust because AI agents access data at machine speed, overwhelming legacy systems. Our new Beyond Zero model shifts security boundaries from applications to individual actions, combining static policies with dynamic, AI-driven reasoning. This approach enables real-time, risk-based authorization for both humans and agents, creating a self-defending enterprise capable of handling millions of decisions per second without disrupting productivity.
The assumptions underpinning BeyondCorp—that accessors are human, that actions occur at human speed, and that applications are the correct boundary for trust—are no longer sufficient.
- skybrian
Maybe this is more interesting in the context of the HuggingFace incident? Suppose you have an AI and some access controls for what it can do. If the access control is too broad or has a bug, can you still detect suspicious activity and slow it down long enough for someone to be alerted?
- kriro
"""
Beyond Zero shifts the trust boundary from the application to the action being performed on a piece of data in realtime—and from after-the-fact investigation to in-the-moment evaluation and containment. It augments BeyondCorp’s foundational identity with a “brain” capable of reasoning about the context and intent of a specific request in realtime.
"""
Doesn't this simply shift the attack vector? Compromising this overlord brain now becomes a new target.
- firasd
Honestly I think non-malicious odd behavior is under-weighted when it comes to AI agents. Even the example in this paper is about someone suspiciously accessing sales data when "why did you do that" often comes down to something in the model's training that fired as a reflex
I wrote about this a few days ago https://firasd.substack.com/p/accidental-data-loss-in-claude...
"Many researchers have made demos along these lines:
An agent is asked to check a webpage like example.com
The webpage asks for a name to proceed further
The agent calls example.com/evil?myname=John, thus sending the user’s name from the context window to the external server.
In practice, however, these elaborate ‘confused deputy’ exfiltration attacks seem rare compared to widely-reported data loss incidents.
The risk of undermining the user’s interests through clumsiness deserves at least as much scrutiny as the risk of leaking secrets."
So while the idea of shifting the permission boundary from the app level to the action level makes sense, what we should also have is some 'failsafes', eg. if the action says 'delete' then maintain a rollback window, if the action is 'send an email' then maintain an events log. Preparing for AI agents means expanding auditability and reversibility in software
- FailMore
If this is interesting to you (as it is to me) but you want to quickly digest it instead of read the paper, Claude + SmallDocs[1] converted it into a slideshow which serves it to you in bitesized ideas:
https://smalldocs.org/s/2SH6FHiUK1mcym24Z8E37I#k=2Sk6c_IdKJL...
[1] I am the developer behind SmallDocs.
- insumanth
I was confused when google acquired "Wiz" for so much money
But now, it makes sense to some extent.
Security is non-negotiable in AI Era