LLM Security

258 reports

All LLM Security Agentic AI Industry News Research Supply Chain Prompt Injection First Look: Security Regulatory Jailbreaks Adversarial ML Data Poisoning Model Theft AI Security Tools Security Operations

Delphi Ships AI Karamo Brown Clone for Kē Wellness App

FIRST LOOK ATLAS OWASP MEDIUM ▲ 6.2 TechCrunch AI Jun 19, 2026

Karamo Brown's Kē wellness app deploys an AI digital clone of the celebrity — voice, persona, and advisory content — built by Delphi from interviews, podcasts, and public clips, enabling real-time conversational coaching at scale. For defenders, celebrity-clone architectures introduce layered risks: the training corpus is largely public and manipulable, the voice synthesis surface is exploitable for deepfake derivation, and the mental-health context creates elevated harm potential if the persona is hijacked or jailbroken. Security teams evaluating similar deployments should treat the persona boundary as a primary control point, since users in vulnerable emotional states are disproportionately exposed to manipulation if guardrails fail.

AWS SageMaker Ships 100+ Inference Metrics to CloudWatch

FIRST LOOK ATLAS OWASP MEDIUM ▲ 6.2 AWS Machine Learning Blog Jun 19, 2026

AWS has released a deep observability layer for SageMaker AI inference endpoints, emitting over 100 metrics covering GPU health, KV cache pressure, token-level latency, and traffic distribution into a native CloudWatch Insights dashboard with PromQL-compatible export. For defenders, this centralised telemetry surface introduces new reconnaissance and exfiltration vectors: an adversary with read access to CloudWatch or connected third-party tools (Grafana, Datadog) can infer model architecture, request patterns, and capacity limits without touching the model itself. The richness of these signals also raises insider-threat risk, as operational staff now have granular visibility into inference behaviour that can be leveraged to reverse-engineer model characteristics or plan targeted denial-of-service campaigns.

AWS Launches Amazon Bedrock AgentCore Harness

FIRST LOOK ATLAS OWASP HIGH ▲ 8.2 AWS Machine Learning Blog Jun 19, 2026

AWS has made Amazon Bedrock AgentCore Harness generally available, providing a managed abstraction layer that reduces agent deployment to two API calls while bundling sandboxed compute, persistent memory, tool gateway, browser access, identity management, and observability. For defenders, this dramatically lowers the barrier to deploying autonomous agents with filesystem access, shell execution, web browsing, and multi-provider model switching — compressing what was a weeks-long infrastructure project into minutes. Security teams face an expanded attack surface where prompt injection, tool abuse, cross-session memory poisoning, and supply chain risks through AWS-curated skill catalogs now arrive as a single, tightly integrated managed service rather than individually reviewable components.

Microsoft AutoGen Studio RCE via MCP Bypass

ATLAS OWASP HIGH ▲ 9.1 Microsoft Security Blog Jun 19, 2026

Researchers at Microsoft identified a three-stage exploit chain in AutoGen Studio that allows a malicious web page visited by a browsing AI agent to reach the host's local Model Context Protocol (MCP) WebSocket and spawn arbitrary processes. The chain exploits a bypassable origin allowlist, authentication middleware that excluded MCP endpoints, and unsanitised URL-derived command parameters. Although the vulnerable surface was never shipped in a PyPI release, the finding exposes a systemic architectural risk in any agent framework that combines untrusted browsing with privileged localhost services.

Orphaned AI Agents Bypass SailPoint Identity Controls

ATLAS OWASP HIGH ▲ 7.2 The Hacker News Jun 19, 2026

Enterprises deploying internal AI agents face a growing identity accountability gap: when the employee who created an autonomous agent leaves, the agent's access tokens and credentials often remain active and unmonitored. Traditional access management tools fail to detect this risk because they treat AI agents as static software rather than identity-bearing entities capable of exfiltrating sensitive data. The problem compounds at scale as shadow AI deployments proliferate across organizations without centralised visibility or ownership tracking.

Anthropic's Mythos 5 and Fable 5 Hit by Export Block

FIRST LOOK ATLAS OWASP HIGH ▲ 7.2 TechCrunch AI Jun 18, 2026

The Trump administration's overnight export block of Anthropic's Mythos 5 and Fable 5 models — triggered by reported safety guardrail bypass vulnerabilities flagged by Amazon — has exposed the fragility of international AI supply chains built on U.S.-controlled infrastructure. For defenders, this event crystallises a critical dependency risk: organisations and governments that have embedded American AI models into critical systems now face the possibility of abrupt, unexplained access revocation with no remediation path. Security teams must now treat AI vendor access continuity as a threat vector equivalent to a third-party SaaS outage, and accelerate contingency planning around model substitution and sovereign alternatives.

AWS Launches Amazon Quick Autonomous Agents

FIRST LOOK ATLAS OWASP HIGH ▲ 8.2 AWS Machine Learning Blog Jun 18, 2026

AWS has shipped autonomous agents in Amazon Quick, an AI assistant that continuously executes tasks — including CRM updates, email drafting, and compliance monitoring — on behalf of users while connected to dozens of enterprise data sources and applications. This dramatically expands the attack surface for business-context compromise: a single successful prompt injection or account takeover can now translate into persistent, automated actions across an organisation's entire connected app ecosystem. Defenders must treat these agents as privileged service accounts with broad, continuous write-access, requiring dedicated monitoring, least-privilege scoping, and explicit human-in-the-loop gates for sensitive actions.

OpenAI Launches ChatGPT for Science with Institutional Access

FIRST LOOK ATLAS OWASP MEDIUM ▲ 5.8 BleepingComputer Jun 18, 2026

OpenAI is internally testing a specialised 'ChatGPT for Science' subscription tier, likely restricted to verified universities and research institutions, building on capabilities from GPT-Rosalind — a purpose-built life sciences model already deployed under a trusted-access structure with select pharma partners. The gated, domain-specific nature of this offering creates novel identity and access verification attack surfaces, as threat actors will likely probe credential and institutional verification mechanisms to gain privileged access to specialised scientific knowledge. Defenders at academic and research institutions should anticipate increased phishing campaigns targeting institutional credentials and prepare governance frameworks for AI use in sensitive research environments.

Z.ai Releases GLM-5.2 Open-Weights 753B LLM

FIRST LOOK ATLAS OWASP HIGH ▲ 6.2 Simon Willison Jun 18, 2026

Z.ai has released GLM-5.2, a 753-billion-parameter mixture-of-experts model under an MIT license, ranking as the top open-weights model on the Artificial Analysis Intelligence Index and second on the Code Arena WebDev leaderboard. For defenders, the combination of frontier-level capability, unrestricted open-weights distribution, and a 1-million-token context window materially lowers the barrier for threat actors to self-host a highly capable model outside any provider's safety controls. The model's agentic coding performance and massive context window expand the viable attack surface for automated code generation, targeted phishing, and large-scale document analysis without API-level monitoring.

CrowdStrike Launches Continuous Identity for AI Agents

FIRST LOOK ATLAS OWASP HIGH ▲ 6.8 CrowdStrike Blog Jun 17, 2026

CrowdStrike's Continuous Identity for AI Agents introduces persistent, trackable identity primitives for agentic workflows — but persistent identities are also persistent targets. Attackers who compromise an agent identity gain a durable, trusted foothold that can persist across sessions and tool invocations without the natural expiry of human session tokens. The feature's integration into the Falcon platform means agent identity tokens, if stolen or forged, may carry elevated detection-suppression trust within the same security toolchain defending the environment.

Anthropic Ships Claude Fable 5 with Exploit Generation

FIRST LOOK ATLAS OWASP CRITICAL ▲ 8.7 Wired Security Jun 17, 2026

Anthropic's Mythos 5 and Claude Fable 5 represent the arrival of frontier AI models with demonstrated, advanced vulnerability discovery and exploit-development capabilities — a capability class that will rapidly proliferate across multiple vendors and open-weight releases. The core attack surface is not model-specific: guardrail bypass of the consumer-facing Fable 5 exposes full Mythos-grade offensive capability to any actor who can defeat the content filters, while the broader proliferation trajectory means defenders must assume adversary access to equivalent capabilities within months. The regulatory response addresses a single vendor while doing nothing to raise the floor for the broader ecosystem of competitive and open-weight models following close behind.

Google Launches Android 17 with Gemini Omni Integration

FIRST LOOK ATLAS OWASP HIGH ▲ 7.2 TechCrunch AI Jun 17, 2026

Android 17 embeds Gemini Omni and multiple AI models (Lyria 3, AudioLM) directly into OS-level functions including video editing, call handling, screen recording, and emergency detection, dramatically expanding the attack surface for AI-assisted exploitation on mobile endpoints. The deep integration of conversational AI with device sensors, media pipelines, and inter-app communication creates novel prompt injection and data exfiltration vectors that existing mobile threat defences were not designed to address. The simultaneous AirDrop interoperability expansion and cross-device Pixel Watch mirroring further widen the lateral movement surface across the Google hardware ecosystem.

NVIDIA Launches XR AI for Agentic AR Glasses

FIRST LOOK ATLAS OWASP HIGH ▲ 8.2 NVIDIA AI Blog Jun 17, 2026

NVIDIA XR AI puts multimodal agentic systems directly into AR glasses, fusing continuous video, audio, depth, and pose data with enterprise knowledge retrieval and tool execution — creating a persistent, always-on sensor exfiltration and prompt injection surface that sits inches from a worker's face. The framework connects to industrial systems, digital twins, and enterprise RAG backends, meaning a compromised agent can pivot from perceptual data into operational technology networks. Because the inputs are environmental and largely uncontrolled, adversarial content placed in the physical world (signage, screens, spoken commands) becomes a viable injection vector against enterprise infrastructure.

Vertex AI SDK Bucket Squatting Flaw Enables Model Hijack

ATLAS OWASP HIGH ▲ 8.5 The Hacker News Jun 17, 2026

A vulnerability in the Google Cloud Vertex AI Python SDK allowed unauthenticated attackers to intercept model uploads by pre-registering predictable staging bucket names — a technique Unit 42 calls 'Pickle in the Middle'. Once a malicious model replaced the legitimate upload, arbitrary code executed inside Google's serving infrastructure via pickle deserialization. Google patched the flaw in v1.148.0 after disclosure in March 2026, but the incident highlights systemic risks in ML pipeline supply chains.

Anthropic Mythos Model Theft: China-Linked Access

ATLAS OWASP CRITICAL ▲ 8.5 The Verge AI Jun 16, 2026

The White House reportedly believes a China-linked group accessed Anthropic's Mythos AI model, prompting export restrictions on the technology. If confirmed, the breach represents a significant national security threat, as adversaries could exploit the model directly or use knowledge distillation to replicate its capabilities. Separately, reports of jailbreak vulnerabilities in Mythos and Fable compound concerns about unauthorised access to frontier AI systems.

LLM Security

Stay ahead of the threat.