30
One in four MCP servers opens AI agent security to code execution risk
Noma Security whitepaper finds most popular AI Skills and many MCP servers carry high-risk capabilities, with state changes most prevalent.
Noma Security analyzed hundreds of popular MCP servers and Skills across eight risk categories, finding most widely used Skills carry at least one risky characteristic and a typical enterprise runs well over a hundred high-risk agent tools, with arbitrary code execution common across MCP servers. The most prevalent risk is the ability to change state or data, and named toxic combinations include ContextCrush data leakage, ForcedLeak via poisoned Salesforce CRM records, DockerDash supply-chain compromise, the Replit production database deletion, and the hijacked Amazon Q VS Code extension. Building on OWASP LLM06:2025, the paper proposes the No Excessive CAP framework of capabilities, autonomy, and permissions, recommending allowlisting, MCP version pinning, approval gates on irreversible actions, and user-scoped expiring credentials.
62
AI Agents Are Here. So Are the Threats.
Unit 42 demonstrates nine framework-agnostic attack scenarios against AI agents built with CrewAI and AutoGen, causing data leakage, credential theft and remote code execution.
Palo Alto Networks Unit 42 investigated how attackers can target agentic applications, implementing two functionally identical apps with the open-source CrewAI and AutoGen frameworks and executing the same attacks on both. Nine attack scenarios produce outcomes including information leakage, credential theft, tool exploitation and remote code execution. Findings show most vulnerabilities are framework-agnostic, arising from insecure design patterns, misconfigurations and unsafe tool integrations rather than flaws in the frameworks themselves. The team published defense strategies per scenario and open-sourced the source code and datasets on GitHub.
58
30
When an Attacker Meets a Group of Agents: Navigating Amazon Bedrock's Multi
Unit 42 red-teamed Amazon Bedrock multi-agent applications, demonstrating prompt-injection attack chains that leak agent instructions and invoke tools, mitigated by Bedrock Guardrails.
Unit 42 red-teamed Amazon Bedrock Agents' multi-agent collaboration in Supervisor and Supervisor with Routing modes. The demonstrated attack chain detects the operating mode, discovers collaborator agents, delivers attacker-controlled payloads, and can disclose agent instructions and tool schemas and invoke tools with attacker-supplied inputs. No vulnerabilities were found in Bedrock itself, and the built-in prompt attack Guardrail blocked the attacks when properly configured. The researchers collaborated with Amazon's security team and frame the findings as a broader prompt injection risk for LLM-based systems.
55
Google’s $10,000 refund test shows why AI agents need zero trust
Google released an open-source zero-trust reference architecture for AI agents defending against prompt injection via signed database writes, gVisor sandboxing, and deterministic gating.
Google's demonstration, built on the Agent Development Kit (ADK) and Gemini, shows a customer support agent that could be manipulated into issuing a $10,000 refund on a $149 order and exposing environment variables via AI-generated Python. The architecture adds three security layers outside the model: cryptographic signatures on database writes verified via Cloud KMS backed by Cloud HSM, gVisor sandboxing of generated code with network egress disabled, and a Semantic Gateway applying deterministic checks to prompts and tool calls. It treats system prompts as insufficient boundaries because prompt injection, prompt tuning, or model updates can change agent behavior.
48
Recursive Code World Models: Building Complex Worlds through Recursive Scene Programs
RCWM reconstructs complex 3D worlds as executable code from a single image using recursive scene programs with global-local-global solver recursion.
The paper introduces Recursive Code World Models, coupling a Recursive Scene Program representation with a recursive construction solver for image-to-3D-world reconstruction. Each solver call establishes the whole scene, recursively reconstructs unresolved parts, and revisits the whole to refine composition, while a vision-language coding agent compares reference images with scene renders to guide refinement. RCWM outperforms prior code-based image-to-scene reconstruction methods, and ablations show deeper recursive calls improve fine-scale reconstruction.
33
10 most critical LLM vulnerabilities
OWASP updated its Top 10 LLM application vulnerabilities, ranking prompt injection first and elevating excessive agency to third amid agentic adoption.
OWASP refreshed its Top 10 list of critical vulnerabilities in LLM applications, for the first time incorporating real-world incident data alongside expert voting. Prompt injection and sensitive information disclosure remain first and second, while excessive agency jumped from sixth to third as agentic systems that call APIs and execute code proliferate. Unbounded consumption of AI resources rose in prominence, while improper output handling dropped to the bottom as output sanitization becomes widespread. The list includes remediation guidance such as strict output schemas, human-in-the-loop approvals, and least-privilege credentials held in application code.
55
30
OpenAI puts major frontier AI training run on hold over cyber risks
OpenAI paused its largest frontier RL training run for two weeks to harden research environments after Astra showed potentially critical cybersecurity capability.
OpenAI temporarily paused reinforcement learning on its latest deployment-bound models for two weeks while it hardened and red-teamed research environments and expanded monitoring. The pause followed the OpenAI-Hugging Face incident and preliminary evidence that the upcoming Astra model may meet the Critical cybersecurity capability threshold in its Preparedness Framework. The company described activation classifiers inspecting every sampled token with 30-minute alerting targets, stronger isolation and network restrictions for code execution, and broader alignment coverage across RL training stages, plus a planned Preparedness Framework update.
72
GPT-6 Astra Scores 100% on ExploitBench as OpenAI Blocks PoC Exploit Requests
OpenAI releases GPT-6 Astra, scoring 100% on ExploitBench, but restricts it to secure code review by blocking PoC exploit generation.
OpenAI officially unveiled GPT-6 Astra days after the model reached the "Critical" cybersecurity capability threshold under its Preparedness Framework. The model claims 100% on ExploitBench (versus 78.5% for GPT-5.6 Sol), 98% on FrontierMath Tier 4, and 99.9% on ARC-AGI-3, and demonstrated exploit development including on two zero-days disclosed between June and August 2026. The released version is limited to secure code review and patching and refuses proof-of-concept exploit requests, with less restrictive safeguards planned via OpenAI Daybreak. OpenAI also launched a $1 billion "Daybreak for Frontline Defenders" program for critical infrastructure sectors and a pilot with the US MS-ISAC for public sector and water system defenders.
92
How to secure edge AI in customer-owned environments
Microsoft outlines security architecture guidance for edge AI, urging runtime attestation, artifact provenance, and deterministic mediation of model actions.
Microsoft details how edge AI shifts trust responsibilities to customers operating their own infrastructure, where prompt injection, model tampering, and malicious firmware updates can occur alongside model weights, credentials, and physical-system access. The guidance recommends verifying runtimes with attestation, verifying AI artifacts with provenance, and constraining model actions through a deterministic mediator outside the model. It also covers new exposure surfaces from MCP, multi-agent systems, and computer-use agents running in disconnected or hostile edge environments.
33
60
Generating running routes with GPT-6 Astra and ChatGPT Work
Simon Willison used ChatGPT Work with GPT-6 Astra to generate 5K/10K running loops from OpenStreetMap data, noting the agent's executed code was not visible.
Simon Willison prompted ChatGPT Work running GPT-6 Astra (Max) to design looping 5K and 10K running routes from his address using OpenStreetMap data. The agent ran for 27 minutes and delivered an embedded visualization plus downloadable GPX and GeoJSON files, stating it used Nominatim for geocoding and Overpass to fetch local OSM roads and trails. He notes frustration that the actual executed code and exact actions were not visible in the ChatGPT UI, highlighting a transparency gap in agentic AI products.
12
Zyxel Releases Patches for Firmware Vulnerabilities in EoL NAS Models
60
OpenAI Launches GPT-5.6-Cyber with Reduced Safeguards for Exploit Development
OpenAI released GPT-5.6-Cyber for vulnerability research and pentesting via Daybreak Red, completing 95% of advanced cyber task evaluations.
GPT-5.6-Cyber, built on GPT-5.6 Sol, targets zero-day discovery, exploit chain development and incident response with reduced refusals, scoring 95.0% on OpenAI's Advanced Cybersecurity Completion Rate versus 1.5% for GPT-5.6 Sol and 57.3% for GPT-5.5-Cyber. The model found CVE-2026-15903 (CVSS 8.8), an out-of-bounds read/write in Chrome's V8 JavaScript engine that Google patched in mid-July 2026. It is available to trusted partners including CrowdStrike, Palo Alto Networks and Cloudflare through the Daybreak Red access tier.
78
30
Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye
METR analysis finds AI accelerating cyber vulnerability discovery, while SPADE self-play environment generation improves Qwen3 reasoning benchmark scores at 30B scale.
Import AI 470 discusses a METR research note reporting differential acceleration from AI: major acceleration in reported cyber vulnerabilities (cURL, OpenSSL, Firefox, Microsoft, NVD, OSV), minor acceleration in mathematics, and no measurable acceleration in AI-research optimization benchmarks. It also covers SPADE, a self-play framework from a multi-university team (University of Washington, Stanford, MIT, CMU, and others) that co-evolves executable training environments and agent capability using Environment Designer and Reasoning Agent roles with hint-based regret rewards. Trained on Qwen3-4B-Instruct-2507, Qwen3-8B, and Qwen3-30B-A3B-Instruct-2507 via GRPO (400 rollouts of 25 environments), SPADE lifted the 30B-A3B game-environment suite average to 58.3, +8.1 over base, and improved tool-use results across backbones. The issue also references Hawkeye for building better GPU kernels.
42
Agent Harness vs Agent Framework vs MCP: Which Layer Owns the Loop, State, Tools, Permissions, and Recovery
Architecture explainer separates agent harnesses, frameworks, and MCP by which layer owns the loop, state, permissions, and recovery.
The article distinguishes agent harnesses (OpenAI Codex, Claude Agent SDK), which own the execution loop, sandbox, permission model, and recovery; frameworks (LangGraph, OpenAI Agents SDK, Microsoft Agent Framework), which supply composable primitives; and MCP, a stateless JSON-RPC wire protocol governed by the Linux Foundation's Agentic AI Foundation since December 2025. An ownership matrix maps the execution loop, state, tool transport, permissions, recovery, sandboxing, and multi-agent orchestration to each layer. The 2026-07-28 MCP specification made the protocol fully stateless, retiring the initialize handshake and session headers.
35
Security leaders must prepare for likely threats, not sensationalized agentic attacks
CSO opinion argues agentic AI attacks mostly exploit mundane vulnerabilities, urging defenders to train on realistic threat profiles rather than sensational containment breaches.
An opinion piece contends recent reports of AI models 'breaching containment' at OpenAI, Anthropic, and Meta overshadow the more likely risk: AI agents exploiting conventional unpatched flaws and insecure APIs. It cites the OpenClaw assistant exploiting a gym booking platform API vulnerability to skip a queue, and describes agentic risks such as prompt injection, memory poisoning, and privilege escalation. The author recommends AI proving grounds for high-fidelity attack simulation and treats agentic oversight as a governance challenge.
30
CodeTD: Topology of Attention Detects Hallucinations in Code LLMs
CodeTD detects hallucinations in code LLMs before execution by analyzing topological patterns of attention maps, outperforming recent baselines.
CodeTD applies topological data analysis (TDA) to code LLM attention maps to quantify prompt-generation mismatch as a pre-execution correctness signal. Experiments cover HumanEval, MBPP, BigCodeBench, and MultiPL-E across 5 programming languages and 10 code LLMs up to 34B parameters. The method outperforms recent baselines and transfers between coding benchmarks, helping catch code that fails the task or embeds security vulnerabilities.
24
30
Google Open-Sources Mantis: A Modular Skills Toolkit That Lets Coding Agents Find, Reproduce and Patch Vulnerabilities
Google open-sourced Mantis, an Apache-2.0 modular skills toolkit that lets AI coding agents find, reproduce, and patch vulnerabilities with sandboxed verification.
Google released Mantis on GitHub under Apache 2.0 as a stack-agnostic set of slash-command skills that chain through the full vulnerability lifecycle: mining version history, building threat models, filtering findings, reproducing bugs in gVisor or network-disabled VMs, assembling exploit chains, patching, and scoring residual risk from 1 to 10. It runs with Gemini CLI, Antigravity CLI, the Google ADK, or comparable agent frameworks, and a supervisor skill (/mantis-meta-agent) can drive the whole loop. Google says the design targets the sub-7 percent true-positive rate of naive AI code scanning, and that its hierarchical summary tree cuts token overhead by over 85 percent. The toolkit is deployable for local and internal evaluation but not yet recommended for production.
55
30
PuzzleMask: Abusing Plain Prose as a Covert AI Attack Vector
Check Point details PuzzleMask, a plain-prose technique that bypasses LLM gatekeeper policy checks, letting hidden payloads reach target models unreviewed.
Check Point Research describes PuzzleMask, a prompt-crafting technique that hides policy-violating payloads inside plain-English prose wrappers, bypassing quick LLM-based policy checks without emojis, Base64, or invisible formatting. The researchers tested 23 automated prompts against gatekeepers including GPT-4o-mini, GPT-OSS-Safeguard 20b, Claude 3 Haiku, and Llama Guard 3, and all were classified as safe despite policies that flagged the plain versions. When submitted to GPT-5 in thinking-high mode with a Python interpreter, the target model extracted and acted on the payload in over 90% of trials. The technique is not itself a jailbreak but can carry a jailbreak prompt as payload; mitigations include input paraphrasing, hardened gatekeeper policies, and output monitoring.
55
30
30
30
30
Ex-Deepmind VP Vinyals says AI self-improvement is coming but won't trigger an intelligence explosion
Ex-DeepMind research VP Oriol Vinyals says recursive self-improvement is coming but slow, and co-founds Discovery Loop with Jeff Dean to automate research.
Oriol Vinyals, former VP of Research at Google DeepMind who worked on AlphaStar, AlphaCode, and Gemini, argued at Agentic AI Summit 2026 that recursive self-improvement will progress gradually without an intelligence explosion. He identifies idea generation ('research taste') and evaluation as the two biggest bottlenecks, noting benchmarks like SWE-Bench Pro and ML-Bench mostly test the already-working steps and suffer from overfitting and scheming. He is co-founding Discovery Loop with Jeff Dean as CEO, Sanjay Ghemawat, and Quoc Le to automate the full research cycle, starting with AI research.
62
How MCP Servers Can Expose Enterprise Secrets
MCP servers holding AI agent credentials risk secret exposure via plaintext configs, credential sprawl, prompt injection, and over-permissioning; mitigations include centralization and least privilege.
The article examines how Model Context Protocol servers, which hold API keys, tokens, and service-account credentials for AI agents, can leak enterprise secrets. Documented exposure paths include plaintext credentials in config files, ungoverned credential sprawl, prompt injection, over-permissioning, and untrusted third-party servers. It cites CVE-2025-6514 in mcp-remote (400,000+ downloads), where a malicious server triggered OS command injection leading to remote code execution. Recommended mitigations include centralized secret stores, short-lived auto-rotated credentials, least privilege, and human approval for sensitive actions.
48
The AI Supply Chain Has a Security Problem, and Much of It Is Sitting on the Open Internet
Researchers counted 36,769 publicly reachable self-hosted AI endpoints, only about 2% behind HTTP authentication, exposing Ollama, vLLM, and Flowise to abuse.
A Mysterium VPN study found 36,769 self-hosted AI endpoints reachable through internet scanning, with only 2.02% returning an HTTP authentication challenge. Open WebUI accounted for 18,529 reachable instances, Ollama for 6,935 fingerprinted hosts, and 5,223 agent-builder and workflow platforms were exposed, often holding API keys, database credentials, and other secrets. The report highlights LLMjacking risk from exposed Ollama APIs, a critical Flowise bug (CVE-2026-40933), leaked n8n tokens, and prior SentinelOne/Censys research finding roughly 175,000 exposed Ollama hosts in 130 countries.
45
We have a year to fix security everywhere
Blog post warns that cheap open-weight GLM 5.3-flash, once abliterated, could enable mass AI-driven vulnerability exploitation, urging industry-wide patching now.
An essay argues that Z.ai's open-weight GLM 5.3-flash—runnable locally on roughly $6k consumer hardware at 20-45 tokens/second—combined with 'abliterated' variants from groups like DeAlignAI that score 0% on HarmBench-320 puts dangerous hacking capability in nearly anyone's hands. GLM 5.3 scores 84.5% on CyberGym and 54.4% on ExploitBench, versus GPT-6 Astra's 100% and GPT-5.6 Sol's 78.5%, and the author cites evidence of frontier models exploiting real-world infrastructure. The author calls for using LLMs (Project Glasswing, Daybreak) to find and fix vulnerabilities industry-wide before adversaries weaponize cheap open models.
55
[AINews] DeepSeek v4.1-Flash: 763B-P8B-D16B novel causal Encoder–Decoder architecture with vision marks the Return of the Whale
DeepSeek released V4.1-Flash, an open-weight 763B-parameter model with a novel causal encoder-decoder architecture, 1M context, vision input, and MIT license.
DeepSeek launched V4.1-Flash, an open-weight MIT-licensed model using a novel causal encoder-decoder architecture with 763B total parameters and asymmetric active parameters: 8B for prefill and 16B for decode. It supports 1M-token context and text+image input, priced at $0.30 per 1M input and $1.20 per 1M output tokens with a 50% off-peak discount. Artificial Analysis scored it 40 on its Intelligence Index, above DeepSeek V4 Pro 0813, and Vals ranked it the #1 open-weight model ahead of Kimi K3. Baseten shipped day-0 support and Ollama began rolling it out to paid subscribers.
80
Claude Mythos Executes End-to-End Intrusion From Initial Access to Full Domain Compromise
Anthropic's Claude Mythos Preview, its most cyber-capable model, autonomously completed an end-to-end enterprise intrusion simulation in restricted-access testing.
Anthropic's April 2026 system card describes Claude Mythos Preview as the first model to solve a private cyber range end to end and finish a corporate-network attack simulation an expert would need 10+ hours to complete. It scored 100% pass@1 on a 35-challenge Cybench subset and 0.83 on CyberGym versus 0.67 for Claude Opus 4.6. The model is limited to vetted partners under Project Glasswing; it failed an OT cyber range and could not find novel exploits in a fully patched sandbox.
75
Quoting Calif Research
Calif Research demos WeWorm, an AI-built zero-click worm that spreads via WeChat calls on iOS and Android without user interaction.
Calif Research released a demo of WeWorm, described as the first zero-click worm to spread through WeChat calls across iOS and Android; victims need not answer the call and hear nothing even if they do. The team used AI to find the underlying bug and write a remote code execution exploit in about two days, then spent one more week building the worm. The researchers argue AI can now do most of the exploit-development work, with humans supplying judgment on targeting and safe testing.
55
EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents
EvoSafeHarness auto-synthesizes per-model, per-domain safety harnesses, cutting prompt-injection attack success on AgentDojo to 0.0% at 82.8% utility.
EvoSafeHarness is an optimization framework that synthesizes deployable safety harnesses for frozen LLM agents in a target domain, jointly searching natural-language policies and executable code logic guided by model behavior, domain specifications, and adversarial review. On DecodingTrust-Agent it reduces average attack success rate from 45.6% to 10.0% at a 3.3-point utility cost, and on AgentDojo reaches 82.8% utility at 0.0% ASR, twice CaMeL's utility at that operating point. It keeps mean ASR below 20% under adaptive PAIR attacks and transfers unchanged to unseen AgentDyn suites. The analysis finds domain semantics determine required safety relations while model and runtime behavior determine enforcement points.
48
Most of the bugs Claude Mythos found have never been checked by a human
Echo's analysis found only 1,900 of 23,019 Claude Mythos-found vulnerabilities were externally reviewed, 90.8% held up, but the model overstated most severities.
Echo analyzed results from Anthropic's Claude Mythos Preview vulnerability sweep across 281 open-source projects, which produced 23,019 candidate vulnerabilities, of which only 1,900 were externally reviewed. Of those, 90.8% held up as real, 1,451 of 1,596 maintainer reports were acknowledged, 97 fixes landed upstream, and 88 became advisories, but 14 of the 27 CVE-assigned severity ratings mismatched independent scoring, mostly overstated. On Anthropic's SpiderMonkey benchmark, Claude Mythos turned known crashes into working code execution exploits in 72.4% of 250 trials, versus below 1% for Claude Opus 4.6. Echo cautions the reviewed sample likely was not randomly drawn, so the accuracy figure may not generalize to the other 21,119 unreviewed candidates.
62
Repeat-After-Me: Black-Box Adaptive Visual Prompt Injection
Researchers unveil Repeat-After-Me, a black-box visual prompt injection achieving over 80% success on Qwen3.6-27B and 47% on GPT-5.5.
Researchers present Repeat-After-Me, a black-box adaptive visual prompt injection that induces frontier VLMs to reveal PII or make malicious tool calls via injected images. It exceeds 80% attack success rate on Qwen3.6-27B and 47% on GPT-5.5 even when the benign user prompt is unrelated and does not authorize the injected task. In a real-world OpenClaw Discord deployment, a minimally injected image can overwrite TOOLS.md, enabling later remote code execution and secret exfiltration.
55
OpenAI Astra Brings Autonomous Zero
OpenAI says Astra is its first model rated Critical for cybersecurity risk, able to autonomously find zero-days and build full exploit chains without human guidance.
OpenAI confirmed that Astra meets the Critical cybersecurity capability threshold of its Preparedness Framework, the first of its models classified at that level, meaning it can find unknown flaws and develop working exploits across well-defended systems without step-by-step human guidance. Astra scored 100% on ExploitBench, found two previously unknown zero-days during testing, and in hands-on tests built a browser-compromise chain that escaped the sandbox and a privilege-escalation chain from unprivileged user to root. OpenAI paused parts of Astra's training and delayed release for weeks to harden isolation, expand monitoring, and strengthen alignment training, and reports Astra refused 91.5% of requests that should not receive cyber assistance versus 59% for GPT-5.6 Sol. Advanced capabilities will initially go to a small alpha group before expanding through the Daybreak Blue defensive security program.
82