Access controls for agents
Temporal is expanding its security capabilities so AI agents can operate safely inside enterprise systems. The company announced that team members from Oso, a New York startup specializing in application authorization, are joining the company. Oso's software product and customer base were acquired by holding company Tiny.
Why it matters
While generative AI models handle reasoning, autonomous agents must interact with internal databases, execute multi-step workflows, and call external APIs without breaching security boundaries. Temporal aims to add Oso's security frameworks to its "durable execution" engine, which maintains application state and execution history across failures, so that an agent's permissions and human approvals stay intact.
What changes
According to Temporal, the Oso team will accelerate security features already in development, with existing platform operations remaining unchanged. VP of product Preeti Somal framed the goal as asking, for each action an agent takes, who it is acting for and what that person is allowed to do, and getting the same answer every time the work is retried.
About the company
Founded in 2019 by Amazon and Microsoft veterans Samar Abbas and Maxim Fateev, Temporal provides open-source orchestration tools used by companies including OpenAI, NVIDIA, Netflix, Snap and JPMorgan Chase. The company reached a $12.55 billion valuation in a $550 million funding round last month.
