cd /news/ai-agents/teleport-beams-trusted-runtime-for-a… · home › topics › ai-agents › article
[ARTICLE · art-167228] src=cloudnativenow.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

Teleport Beams Trusted Runtime for AI Agents Reaches General Availability

Teleport announced the general availability of Teleport Beams, a trusted runtime for securing AI agents in production infrastructure, at KubeCon + CloudNativeCon North America 2026 in Salt Lake City; the product had been in public beta since March 2026. Beams gives each agent a delegated identity with zero standing privileges, routes execution plans through human or LLM-judge review, authenticates via IAM, Kerberos and X.509, captures traces through an LLM gateway, scores sessions against the MITRE ATT&CK framework, and runs agents in ephemeral micro-VMs with controlled egress, ingress and DNS. "Most enterprises have not connected agents to their production infrastructure," said Ev Kontsevoy, CEO and co-founder of Teleport, whose Infrastructure Identity Platform the company said is in production at more than 700 enterprises.

read2 min views1 publishedNov 10, 2026

SALT LAKE CITY -- Teleport announced during KubeCon + CloudNativeCon North America 2026 the general availability of Teleport Beams, a trusted runtime for securing AI agents in production infrastructure. Beams had been in public beta since March 2026 and connects agents to cloud accounts, Kubernetes clusters, PostgreSQL databases, MCP servers and other resources without placing secrets or passwords in the agent’s environment.

Four Layers of Control for Agents #

Beams governs agent behavior through four layers of control. Each runtime carries a delegated identity from a user or another agent, and agents start with zero standing privileges, proposing execution plans and access requests that a human or an LLM judge reviews before a separate runtime executes them with only the approved roles. Beams discovers available resources automatically and authenticates through IAM, Kerberos, X.509 and other identity-based methods, while an LLM gateway handles model selection and captures traces of everything an agent does. To control agent swarms, Beams detects goal hijacking and emergent coordination between agents and scores each session against the MITRE ATT&CK framework. Agents run in ephemeral micro-VMs with virtualized networking that controls egress, ingress and DNS.

Additional features include natural-language Agentic Classifiers with seven prebuilt rules, such as flags for exposed secrets, destructive actions and privilege escalation, plus support for custom rules. A Phases capability snapshots the artifacts and memory of an expiring session so work can continue or fork into parallel workflows, and a kill switch terminates an agent’s activity and all of its connections to production infrastructure at once. Beams is built on the Teleport Infrastructure Identity Platform, which the company said is in production at more than 700 enterprises.

“Most enterprises have not connected agents to their production infrastructure. They are blocked by a lack of delegated identity, strong access control without secrets and with full audit and traceability of intent to action,” said Ev Kontsevoy, CEO and co-founder of Teleport. “When the cost of an error is high, traditional sandboxes don’t cut it.” Teleport Beams is generally available now.

── more in #ai-agents 4 stories · sorted by recency
── more on @teleport 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/teleport-beams-trust…] indexed:0 read:2min 2026-11-10 · —