1Password Buys Apono: Access Control for AI Agents Inside Corporate Systems

1Password has acquired Israeli startup Apono, a platform for dynamic access management for people, machines and AI agents inside corporate systems. The deal closes one of the key gaps in modern enterprise security: who controls what an autonomous AI agent does once it is already inside a company's infrastructure.

Why 1Password needs Apono

Traditional identity and access management (IAM) systems were designed for people and service accounts. AI agents are a fundamentally different kind of entity: they act autonomously, request access in real time, and can scale to hundreds of parallel instances within seconds. Static role models with "reader" and "editor" permissions are not built for that kind of dynamism.

Apono targets exactly this problem. The platform determines in real time what each participant, whether a human, a machine or an AI agent, is allowed to do inside company systems. It grants exactly the permissions needed right now, and only for as long as necessary, then revokes them automatically. This is the principle of least privilege, rethought for the agentic era.

The startup was founded in Tel Aviv. Before AI agents appeared, Apono specialized in privileged access management for DevOps teams and cloud infrastructure, an area where temporary permissions on production systems are critical. The Israeli business publication Calcalist was the first to report the deal. The purchase price has not been disclosed.

A new attack surface

The agentic AI market is growing faster than security infrastructure can adapt. Companies are deploying autonomous agents, LLM pipelines and Copilot systems, and each of them potentially has access to databases, internal APIs and corporate services. Without control over what an agent does after getting inside the perimeter, traditional defenses lose their point.

"Agentic AI creates a fundamentally new attack surface: not people who

can be trained to be careful, but autonomous systems that act on their own" — a typical assessment from enterprise security specialists.

Here is what a solution of Apono's class offers corporations:

• Dynamic access policies instead of static roles
• Granular control at the level of each agent and each task
• Full audit of all actions inside the infrastructure
• Automatic revocation of permissions when a session ends or anomalies are detected
• A ready foundation for compliance in regulated industries: finance, healthcare and the public sector

1Password's strategy

The Toronto-based company started as a password manager for business and over the past decade has grown into a notable player in enterprise identity management. Over the last two years, 1Password has steadily widened its scope: from protecting developer secrets to managing machine credentials, and now to controlling AI agents.

The Apono acquisition fits that logic. AI agents are not just a new type of service account: they are a new class of actors with unpredictable behavior, broad capabilities and a minimal set of control standards. Whoever offers the market a clear tool to manage them first takes a position with high switching costs, since enterprise companies very rarely change IAM systems after deployment.

Integrating Apono will give 1Password a single control layer: people, machines and AI agents under one policy umbrella. For chief information security officers (CISOs), this answers a question that most enterprise companies have not yet resolved.

What this means

AI agents already work inside corporate systems, but the tools to manage them are only starting to take shape. 1Password is betting that governance of agentic AI will become the next big market in enterprise security, and buying Apono is an attempt to claim a winning position while competitors are still formulating their strategy.