AppViewX built and deployed a classifier to automatically identify and block model attempts to probe or escape a testing environment or obtain internet access.
Public source
Publisher name
Public post
Interesting take on "Defense in depth for agents" by Anthropic from the post https://lnkd.in/e6maXKFS Some good tidbits in the news post: "While evaluations were paused,…
Company
AppViewX
The Leader in Machine and Agent Identity Management
- Industry
- Computer and Network Security
- Location
- New York, US
- Company size
- 501–1,000 employees
About AppViewX
AppViewX is trusted by the world’s leading enterprises to prevent outages, reduce risk, and ensure compliance through automated Certificate Lifecycle Management (CLM) and PKI solutions. Our platform delivers end-to-end visibility, automation, and policy control while enabling crypto-agility. Today, mid-size enterprises to Fortune 500 companies—including 6 of the top 10 global banks and 5 of the top 10 healthcare providers—rely on AppViewX as the foundation for their digital trust.
See moreLatest activity
Latest activity from AppViewX
12 signals
Presence & Recognition
AppViewX posted a wrap on the
Research & Knowledge
AppViewX published a blog titled How Machine Identity Management Bolsters IoT Security.
Products & Services
AppViewX released a new feature allowing rogue AI Agents to be inserted into roles and situations portrayed by humans, with runtime control different from Human Identities.
Discover more
Similar signals
Similar public activity from other companies.
Technology & Infrastructure
Delinea
Delinea extended the controls already on the table for AI agents, allowing runtime authorisation to evaluate the agent at the moment it asks for access against the policy already running and scopes the credential to that task without standing privilege.
Technology & Infrastructure
Stream.Security
Stream.Security developed production security to find and remove AI agents from a programmer wiki site.
Technology & Infrastructure
Workloft.ai
Workloft.ai built an autonomous fleet that prevents executable content from being fetched from the web and stops any code touching downloaded files at a human, with tools in code as an allowlist.
Products & Services
Sweet Security
Sweet Security enables AI security at scale by discovering approved and shadow AI directly from runtime activity, connecting agents to their identities, tools, data and cloud infrastructure, allowing, auditing, or blocking AI-related behavior as it occurs, and applying policies across clusters, namespaces, workloads, and cloud accounts.
Technology & Infrastructure
Kai