{"slug": "the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety", "title": "The Containment Gap: How Deployed Agentic AI Frameworks Fail Public-Facing Safety Requirements", "summary": "A new audit of three dominant agentic AI frameworks—LangChain, AutoGPT, and OpenAI Agents SDK—found that none provide native compliance with six essential safety principles for public-facing deployments. In a simulated government benefits agent built on LangChain, a single memory-poisoning attack increased wrongful denial rates for targeted applicants to 88.9% while preserving aggregate accuracy, making the corruption difficult to detect. The researchers introduced two lightweight containment mechanisms that eliminated both attack vectors with sub-millisecond overhead, concluding the current framework ecosystem may not meet secure-by-default expectations for high-stakes applications like government services and healthcare triage.", "body_md": "arXiv:2606.12797v1 Announce Type: new\nAbstract: Agentic large language model systems that autonomously invoke tools, maintain persistent memory, and execute multi-step plans are increasingly deployed in public-facing domains, including government services, healthcare triage, and financial advising. We ask whether the frameworks used to build these systems provide architectural-level structural safety guarantees. Applying six containment principles derived from a compositional model of agentic architectures, we audit three dominant frameworks (LangChain, AutoGPT, and OpenAI Agents SDK) and find no native compliance in any of them. Memory integrity, a defense against one of the most prevalent vulnerability classes, is not observed in any of the three evaluated frameworks. We validate these findings empirically: in a simulated government benefits agent built on LangChain, a single memory-poisoning write induces persistent targeted corruption across all tested seeds and backends, increasing the wrongful denial rate for targeted applicants to 88.9%. Under a complex five-factor policy, the same attack preserves aggregate accuracy while increasing targeted wrongful denials by 3.5x, rendering the corruption difficult to detect through standard monitoring. We then introduce two lightweight containment mechanisms: a memory integrity validator and a policy gate, which eliminate both attack vectors with sub-millisecond overhead (<0.2ms per call). We conclude that the current agentic framework ecosystem may not yet meet secure-by-default expectations for public-facing deployments and outline priority architectural interventions to enable trustworthy deployment in high-stakes, socially impactful applications.", "url": "https://wpnews.pro/news/the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety", "canonical_source": "https://arxiv.org/abs/2606.12797", "published_at": "2026-06-12 04:00:00+00:00", "updated_at": "2026-06-12 04:53:23.100524+00:00", "lang": "en", "topics": ["ai-safety", "ai-agents", "large-language-models", "ai-research", "ai-policy"], "entities": ["LangChain", "AutoGPT", "OpenAI Agents SDK", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety", "markdown": "https://wpnews.pro/news/the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety.md", "text": "https://wpnews.pro/news/the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety.txt", "jsonld": "https://wpnews.pro/news/the-containment-gap-how-deployed-agentic-ai-frameworks-fail-public-facing-safety.jsonld"}}