{"version":1,"type":"story","url":"https://digestai.news/story/benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore","json":"https://digestai.news/story/benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore.json","markdown":"https://digestai.news/story/benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore.md","slug":"benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore","headline":"Benchling secures multi-tenant AI agents with Amazon Bedrock AgentCore","summary":"Benchling runs AI‑generated scientific code for thousands of life‑science tenants using Amazon Bedrock AgentCore Code Interpreter in VPC mode. The architecture processes more than 600 code execution sessions per day and serves over 250 distinct tenants each week, all while reporting zero security incidents since its early‑April 2026 launch.\n\nThe solution isolates untrusted code in a dedicated AWS account separate from Benchling’s production environment. Network isolation is enforced with a VPC that has no internet or NAT gateway, a dedicated security group, Route 53 Resolver DNS Firewall (deny‑list, allow‑list, catch‑all), VPC endpoint policies for S3 access, and NACLs that limit traffic to port 443. Per‑job credentials are injected via AWS STS, ensuring each session can only reach the specific tenant’s data bucket. Continuous integration tests simulate DNS tunneling, unauthorized endpoint connections, and out‑of‑scope S3 requests to validate the defenses.\n\nBenchling’s security team chose this managed approach over building a custom sandbox, citing reduced engineering effort and the ability to focus on product security. Jeremy Stashewsky, Application Security Engineer at Benchling, said the pattern lets them “buy instead of build” a secure, scalable AI execution environment.","keyPoints":["Benchling runs >600 AI code sessions daily for >250 tenants weekly using AgentCore in VPC mode.","Architecture isolates untrusted code in a separate AWS account with DNS Firewall, VPC endpoints, and per‑job STS credentials.","Since early April 2026, Benchling reports zero security incidents and no cross‑tenant data leakage."],"whyItMatters":"Shows a practical, scalable way for SaaS platforms to run multi‑tenant AI code safely, setting a security blueprint for the industry.","category":{"slug":"enterprise","name":"Enterprise & Industry","url":"https://digestai.news/category/enterprise"},"entities":{"companies":["Benchling","Amazon"],"models":[],"people":["Jeremy Stashewsky"]},"firstPublishedAt":"2026-09-21T16:27:34Z","updatedAt":"2026-09-21T16:27:34Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"AWS Machine Learning Blog","title":"How Benchling secured multi-tenant AI agents with Amazon Bedrock AgentCore","url":"https://aws.amazon.com/blogs/machine-learning/how-benchling-secured-multi-tenant-ai-agents-with-amazon-bedrock-agentcore","publishedAt":"2026-09-21T16:27:34Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"Amazon Bedrock Powers Enterprise AI Agent Rollout","url":"https://digestai.news/thread/abnormal-ai-uses-aws-bedrock-for-email-threat-detection","storyCount":2},"cite":{"text":"Digest AI, \"Benchling secures multi-tenant AI agents with Amazon Bedrock AgentCore\", 21 September 2026, https://digestai.news/story/benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore","publisher":"Digest AI","title":"Benchling secures multi-tenant AI agents with Amazon Bedrock AgentCore","datePublished":"2026-09-21T16:27:34Z","url":"https://digestai.news/story/benchling-secures-multi-tenant-ai-agents-with-amazon-bedrock-agentcore"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}