As AI agents start to take over business processes that have typically been the responsibility of humans, many of those agents will have to sign in to multiple systems to complete their tasks securely. To help enterprises scalably manage that challenge to modern credential management best practices, 1Password — a company widely known for its password management solution — has announced the addition of agentic AI security capabilities to its Extended Access Management Platform (XAM).
During the past year, there’s been lots of talk about AI potentially taking over many jobs. Bill Gates recently predicted that only three jobs will survive: biologists, energy experts, and the coders of AI itself (he also told Jimmy Fallon that we won’t want to watch computers play baseball). However, given the extent to which most humans have to log in to multiple systems to get their jobs done — sometimes even for just one task — who will manage the credentials securely for those AI agents as they start to proliferate?
Also: The best VPN services (and how to choose the right one for you)
Take software development as an example. As ZDNET’s Joe McKendrick recently reported, OpenAI’s forthcoming Agentic Software Engineer is not just augmenting the current software engineers in your workforce, but is an agentic software engineer that can take a pull request and build an app for you.
The enterprise software development process involves a carefully choreographed workflow that often requires developers to authenticate with multiple solutions across an organization’s continuous integration/continuous development (CI/CD) infrastructure. The idea that agentic software engineers could start to take over the role of actual software engineers could be massively disruptive to the underlying security controls.
This concern only scratches the surface of the challenge. Given the API-driven composability of many enterprise automation systems, it’s also highly likely that AI-agent-directed workflows will need to log in to multiple internal and external services to complete their tasks.
Also: Why scaling agentic AI is a marathon, not a sprint
“We’re standing at the precipice of a massive transformation,” 1Password CEO David Faugno told ZDNET. “AI agents have the potential to be a massive accelerator of productivity and innovation. However, these AI agents will need to be able to execute human-like activities, such as logging in to systems, to perform their work. Unfortunately, legacy identity and access management solutions weren’t designed to govern or secure non-human, non-machine identities like AI agents.”
<!–>
The disruptive promise of AI agents, according to a prepared statement by 1Password co-CEO Jeff Shiner, is that they’ll “operate non-stop at machine speed, scaling workflows in ways we’ve never seen.” Shiner noted that older security models designed for humans can only underserve agentic AI’s potential for speed and scale.
When asked if the new functionality in 1Password’s XAM platform will take passkeys into account, a company spokesperson said: “We’re always working to improve our offerings, but we don’t comment on potential or unannounced products.”
Also: Will AI replace software engineers? It depends on who you ask
With the announcement, 1Password has introduced three new agentic AI capabilities to XAM: partitioned vaults for dealing with AI agent-specific secrets, scoped APIs to enable the programmatic management of vault items into AI workflows, and a software development kit (SDK) to help software engineers with the development and automation of those workflows.
The new functionality will be available to developers through SDKs for Javascript, Python, and Go, and the documentation can be found on 1Password’s official developer portal. XAM’s new agentic AI capabilities are available to customers of 1Password’s Enterprise Password Manager business plan at no additional cost.
–>
Source: Robotics - zdnet.com
