According to a recent LinkedIn post from Basis Theory, the company is introducing a new platform designed to let software agents perform tasks that require sensitive credentials without directly accessing underlying secrets. The post describes a model in which original payment cards, passwords, or API keys remain with the customer while agents receive scoped, revocable credentials tailored to a single job and destination.
The company’s LinkedIn post highlights several initial components, including Agentic Payments that allow agents to complete transactions without seeing card data and PCI Agent Runtimes that can execute existing agent code in a PCI Level 1 environment. It also references a Browser Proxy offering that enables browser automation to handle checkout flows while keeping card numbers out of the developer’s code path.
As shared in the post, Basis Theory is also developing upcoming capabilities such as Browser Aperture for controlled shopping and checkout, and an Agent Credential Proxy to let agents call APIs without holding keys. The post further mentions plans for Managed Authentication and a PCI-compliant LLM Router, suggesting a roadmap aimed at securing emerging AI and automation workflows that rely on third-party agents.
For investors, the content suggests Basis Theory is positioning itself at the intersection of payments security, AI agents, and compliance-sensitive infrastructure. If these tools gain traction among enterprises building automated agents, the company could benefit from growing demand for secure credential management and PCI-compliant environments, potentially strengthening its competitive position in the payment security and developer tooling market.