Credentials are now spread across developer tools, cloud workloads, and AI systems. That makes them harder to track, rotate, and secure. MSSPs and security teams are now being asked to help customers manage credentials for people, software pipelines, cloud workloads, and AI agents without leaving secrets scattered across the business.
1Password is addressing that issue with
1Password Credential Broker, a new product that gives approved users, tools, and workflows access to credentials when they need them. The launch moves 1Password beyond storing secrets in a vault. Credential Broker is designed to release approved credentials only at the time of use.
A response to secret sprawl
Credentials often get copied into repositories, configuration files, environment variables, CI/CD tools, service accounts, and scripts. Once that happens, security teams have a harder time knowing where a credential lives, who or what used it, whether it was rotated, and whether it is still needed.
The problem is growing as companies use more automation. People still need passwords, passkeys, and access tokens, but they are no longer the only ones requesting access. Build pipelines, cloud workloads, SaaS integrations, and AI agents also need credentials to do their jobs.
Jeff Malnick, VP, GM of Developer and AI at 1Password, told MSSP Alert that many organizations still manage human credentials and machine credentials in separate systems, with separate policies, logs, and audit trails. That model becomes harder to maintain as CI/CD pipelines, machine workloads, and AI agents begin accessing sensitive systems autonomously and at machine speed, he said.
“What makes 1Password Credential Broker different is that it does not just store or rotate credentials,” Malnick said. “It brokers access at the exact moment a credential is needed, and no longer.”
Instead of moving a credential into a pipeline or service account and leaving it there, Credential Broker keeps the credential in 1Password, verifies the workload making the request, and releases only the credential approved for that job, Malnick said.
Identity signals move into developer workflows
The first use case is GitHub Actions, where a workflow can verify itself with 1Password before a credential is delivered. CI/CD pipelines are a common place for credentials to get exposed. Build systems often need access to cloud services, registries, deployment tools, and internal systems. When those secrets are stored inside pipelines or copied across environments, they become harder to track, audit, and rotate.
1Password’s approach uses workload identity signals to decide whether a workflow should get a credential. That does not remove every pipeline security risk, but it can reduce the number of static secrets sitting in plain text or loosely managed places.
The company is also positioning Credential Broker as part of a broader identity strategy for people, machines, and AI agents. That connects to its Unified Access strategy, which is focused on discovering, securing and auditing access across human and AI agent activity.
Malnick said the product brings machine and AI agent credentials into the same trusted vault that organizations already use to manage human credentials. “The same governance and audit trail that already covers every human credential now secures the pipelines and agents working on behalf of developers,” he said.
Auditability becomes part of the access model
One important piece for MSSPs is logging. Credential Broker is designed to create an audit trail of credential requests and delivery events, including identity context. That kind of record can help security teams answer basic but important questions: what requested the credential, what was delivered, under what trust relationship, and when the request happened.
Malnick said security teams need more than a record of who requested access. They also need the context behind the request.
“With Credential Broker, access requests are evaluated against the authorization that was granted, rather than simply accepting a valid credential,” Malnick said.
Every credential delivery is logged with attribution that includes the workflow that requested access, the credential it received, and the context behind the request, he said. That visibility becomes more important as machine workloads and AI agents take on more work across enterprise environments.
“As AI agents and machine workloads take on more work, security teams need that record to understand what accessed credentials and whether access matched policy,” Malnick said. “With that record living in 1Password, security teams can review machine and human access in tandem instead of chasing logs across separate systems.”
Where Credential Broker fits in the security stack
Credential Broker is not being positioned as a replacement for privileged access management, secrets management, cloud infrastructure entitlement management, or MDR platforms. Instead, 1Password sees it as a control layer for reducing the number of places credentials need to live.
“Credential Broker is designed to sit alongside secrets management and PAM tools, not replace them,” Malnick said. “Traditional secrets managers help organizations store, rotate, and manage machine secrets. Credential Broker addresses a different problem of keeping credentials entirely out of the systems that consume them.”
Malnick also said the product gives MSSPs a clearer record of which machine workloads and agents accessed which credentials and under what policy. He also pointed to 1Password’s Apono access governance capabilities as part of the company’s broader direction.
“As 1Password expands its access governance capabilities, including through Apono, it will help customers understand where credentials are stored and control how access is used after it's granted,” Malnick said.
Credential management is moving beyond storing secrets safely. Security teams now need to control how credentials are requested, delivered, and audited across people, workloads, and AI agents. For MSSPs and security-focused partners, that creates room for advisory, implementation, and managed services around secrets reduction, CI/CD security, AI access governance, and identity security operations.
