Delinea today extended the reach of its platform for securing identities and credentials to now provide support for artificial intelligence (AI) agents.
Phil Calvin, chief product officer for Delinea, said AI agents are rapidly emerging as a new type of machine identity that, given the privileges being granted to them, represent a rich new target for cybercriminals.
Most cybersecurity teams are already facing significant challenges securing machine identities. Recent Delinea Research suggests that, on average, there are 46 human identities for every human identity. AI agents will soon exponentially increase the number of non-human identities that cybersecurity teams will be expected to secure, he added.
The latest additions to the Delinea portfolio include a Vault AI platform for managing AI credentials, including rotating passwords, and a Secure AI offering that ensures least privilege access controls are being enforced.
In the second quarter, Delinea will also preview Discover AI, a tool that will enable cybersecurity teams to identify unsanctioned use of AI. In the second half of this year, Delinea will also preview an authorization tool that makes use of AI agents to automatically enforce least privilege access policies in real time.
Finally, the company is also planning to roll out at an unspecified date Identity AI, a large language model (LLM) that it is training to enable cybersecurity teams to use natural language to manage privileged accounts.
These offerings will in addition to leveraging the core identity management capabilities originally developed by Delinea will also make it simpler for cybersecurity teams to leverage a governance framework for identities that the company gained with the acquisition of FastPath and a set of cloud infrastructure entitlement management (CIEM) and identity threat detection and response technologies added with the acquisition of Authomize, said Calvin.
It’s not clear how rapidly AI agents are being built and deployed, but there will soon be thousands of them distributed across IT environments. Each one of them will have a unique set of credentials that enables them to access applications and services. Cybercriminals will, naturally, seek to compromise the credentials as part of any campaign to gain access to IT environments. If history is any guide, securing those AI agents will once again be an afterthought unless cybersecurity teams start extending the reach of their identity access management (IAM) tools starting now.
Eventually, the scope of the workflows that can be automated via the orchestration of multiple AI agents is only going to become more complex. As the use of AI agents expands and evolves, the number of mission-critical applications and services they will be dynamically accessing is only going to increase.
AI agents, for all intents and purposes, are a form of digital labor that needs to be provided with secure access to a range of applications and services. In effect, it will be as if an organization suddenly added hundreds of employees to its organization, each of whom needs a set of credentials and an assigned set of privileges. As always, it will be cybersecurity teams that will be held accountable for any cybersecurity incident involving those agents. The challenge and the opportunity now is to get ahead of the rise of AI agents before a cybersecurity team is inevitably overwhelmed by them.
