Oasis unlocks lifecycle management and security for non-human identities and AI agents.
Joined January 2024
- Tweets241
- Following52
- Followers196
- Likes76
Happy Halloween. 🎃
Three real monsters hiding in your environment.
Swipe 👇
#NonHumanIdentity #Halloween
“Is this workload who it says it is?” It’s harder to answer than it sounds.
Humans have faces, fingerprints, and phones. Non-human identities have no biometric, so MFA doesn’t really apply, and what’s left is a secret.
#NonHumanIdentity #IdentitySecurity
An attacker only has to find one to sign in, move around, and leak data.
The fix starts with knowing where every secret lives and who owns it. Adam Ochayon, VP Strategy, Identity, explains. See how Cyera Identity helps cyera.com/platform/identity
Most AI agents only need access for one task, and the task takes seconds.
The risk starts when that access outlives the task. Permissions that stay open indefinitely, or reach sensitive data the agent never needed, can turn a quick job into damage that lasts a lot longer.
Cyera Identity scopes an agent’s access to its task and expires it when the task is complete.
See how to keep permissions tied to the work 👇
cyera.com/platform/identity
#1
We all know someone who’s already connected a new AI chatbot to Salesforce. It saves time. But a stolen OAuth token could let an attacker export customer records.
Here are three questions to ask before giving an AI agent access to your data.👇
#2
Or the support team’s AI agent, which a malicious instruction hidden in a ticket tricks into exposing customer data. Both use access granted for a legitimate purpose.
Before granting it, we need to know what each identity can reach and do. 👇
cyera.com/platform/identity
There’s a lot to gain from bringing data and identity together that doesn’t get talked about enough. Before we dive deeper, let’s break down what each brings to the picture and why they work better together. 👇
Every “dot” is a new privileged insider.
OpenAI’s dots are always-on AI agents with their own desktop, their own browser, and every connector their owner has enabled.
A dot isn’t a feature. It’s an identity, and it needs to be treated like one: inventoried, scoped to its own credentials, granted the minimum access, and kept away from sensitive data it has no reason to reach.
You can’t protect what you can’t locate. Find your sensitive data before the agent does.
Full research from Yaakov (Fufu) Shpigelman:
cyera.com/research/dots-a-do…
#DataSecurity #AIAgents #IdentitySecurity
96% of the permissions workers have are never used.
Cyera's research across 2.4 million workers found that people use just 4% of the access available to them.
Teams need to know what an agent is trying to do, what data is involved, and whether that access is appropriate right now. Identity and data have to meet at the access decision.
Our latest blog by Amit Zimerman and Jason Clark explores why.
Link 👇
cyera.com/blog/identity-meet…
#AgenticAI #IdentitySecurity #AISecurity
If we had to describe an Oasis exit party, it would probably look like this.
Back to the desert. The people who built it, the people who believed in it, and a few stories that are better left off X.
This video captures it better than we can. 👇
Before joining forces with @cyera_io , we took one night to celebrate what this group built together.
Thank you to our team, former teammates, investors, customers, and everyone who believed in Oasis along the way.