Mintive raises $7 million to give AI agents IDs, backed by Karpathy, Dean

Mintive, a US start-up that builds identity and access controls for artificial intelligence (AI) agents used inside companies, announced a $7 million seed funding round on Thursday, October 29, 2026. The round was led by Cowboy Ventures, with Coatue, Operator Collective and Hustle Fund taking part, and brings the company's total funding to $11 million. Its individual backers include Jeff Dean, Google's former chief scientist; Andrej Karpathy, a co-founder of OpenAI; and Frederic Kerrest, a co-founder of the identity company Okta.

An AI agent is software that does not just answer questions but takes actions: it can read and send emails, update customer records, file tickets, move files or run code, often across many company systems. As businesses hand more tasks to such agents, a basic security question arises. Each human employee has a login, a set of permissions and an audit trail. Agents often do not. They may run on a shared key or borrow a person's access, which makes it hard to know which agent did what, or to stop one that misbehaves.

That is the gap Mintive says it fills. Identity and access management, deciding who can use which system and keeping a record of it, is a mature business for human users, led by companies such as Okta. Giving agents their own identities, limits and logs is a newer problem. The company's product, by its own description, lets IT teams decide which cloud services each agent may use, routes agents' requests to AI models through a single gateway, a checkpoint where traffic can be inspected and controlled, and monitors their activity.

The investor list explains some of the attention. Backing from Jeff Dean and Andrej Karpathy, two of the best-known figures in AI research, and from an Okta co-founder signals that experienced people see agent security as a real need. The company's customer claims, which have not been independently verified, suggest early demand from firms in health, finance, education and AI.

For Indian founders and engineers, the opportunity is direct. Indian software and security teams that build or run enterprise systems for clients at home and abroad will increasingly be asked how AI agents are controlled, which creates room for new tools and services. As Indian firms deploy AI agents, data-protection rules that require accountability for how personal data is handled will make such controls more important, not less.

Short-term: Mintive gains money to hire and expand its product, and the round adds to investor interest in security and control tools for AI agents.

Long-term: If AI agents become routine inside companies, giving each one an identity, limits and an audit trail could become a standard requirement, creating a market similar to identity management for human users.

Who is affected: Companies deploying AI agents, IT and security teams, identity and cloud-security vendors, start-up founders building enterprise AI tools, and regulators concerned with accountability for automated systems.

Questions and Answers