Teleport Beams launches to secure AI agents in production
Teleport has made its runtime for AI agents, Teleport Beams, generally available. The product provides enterprises with a way to deploy AI agents into production infrastructure while addressing security and identity gaps that had previously restricted their use.
Editor, Lazyfounder
TABLE OF CONTENTS
Beams addresses AI agent security gaps
Teleport has launched Beams, a runtime for artificial intelligence agents, making it generally available for enterprises to deploy AI agents into production infrastructure. The product targets security and identity gaps that have limited AI agent adoption in enterprise environments.
Ephemeral micro-VMs and delegated identity
Beams provides each AI agent with its own ephemeral micro-virtual machine, called a Beam. These micro-VMs operate with delegated identity, ensuring agents start with no standing privileges. This setup prevents unauthorized access and limits potential damage if an agent is compromised.
Execution plans and access controls
Before execution, Beams requires an approved plan detailing roles and steps. Reviewers—either human or a large language model—can revise this plan to align with security policies. Passwords never enter the agent’s environment, reducing exposure to credential theft.
Audit trails, risk scoring, and kill switches
Every action taken by an AI agent is traced and logged, creating a complete audit trail. Beams scores each session against the MITRE ATT&CK framework to generate a risk summary. Teams can add custom detection rules in natural language to identify specific threats, such as privilege escalation.
The system also supports parallel testing by forking Beams. If an agent behaves unexpectedly, a kill switch can terminate its activity and sever all connections to production systems instantly.
Market adoption and availability
Teleport’s broader Infrastructure Identity Platform is already used by over 700 enterprises, including Nasdaq, DoorDash, and Elastic. Cisco Systems became Teleport’s largest strategic investor in August, signaling confidence in the company’s approach to AI agent security.
Beams was first demonstrated in March, following the publication of the Teleport Agentic Identity Framework. After a public beta, it is now available at beams.run.
How Beams connects to broader AI adoption trends
In August, Cisco Systems announced it had become Teleport’s largest strategic investor, linking Beams to Cisco’s push for intelligence-led services in its partner ecosystem. Earlier Lazyfounder reporting showed Cisco is repositioning its partner programs to focus on AI-driven infrastructure modernization, making Beams a potential default choice for Cisco’s enterprise customers.
A separate Lazyfounder analysis found that AI investments have delivered financial payoffs only for a small group of tech giants. By addressing auditability and least-privilege execution, Beams removes two of the biggest barriers to broader enterprise adoption, potentially expanding the market beyond early adopters.
What this means
Lazyfounder analysis — our interpretation, not reported fact.
Teleport Beams shifts the risk calculation for enterprises adopting AI agents. Earlier background reporting showed most firms see little financial payoff from AI investments, largely because sandboxes and VMs could not guarantee auditability or least-privilege execution. Beams directly answers those pain points: every agent runs in its own ephemeral micro-VM with delegated identity, a full MITRE-ATT&CK risk score, and an immutable audit log. For operators, this means AI agents can now touch regulated production systems without requiring a separate compliance exception for each deployment. Cisco’s August investment also suggests Teleport’s approach is becoming the template for partnership-led AI infrastructure sales.
Key takeaways
- Beams removes the main production blocker for AI agents by providing delegated identity and end-to-end audit trails inside ephemeral micro-VMs.
- Every agent session is scored against the MITRE ATT&CK framework, giving security teams a concrete risk metric instead of vague assurances.
- Teleport’s installed base—over 700 enterprises—and Cisco’s investment make Beams a de facto benchmark for how AI agents will connect to regulated infrastructure.
Related on Lazyfounder
Sources
- SiliconANGLE · 2026-11-10
Teleport makes Beams generally available to bring AI agents into production
This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.
About the author
Editor, Lazyfounder
Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.
More stories by Tarun MottliaGet the LazyFounder Brief
Startup, funding and AI news in a five-minute read. Join the early-access list.
