All news in category “AI and Security Pulse”

🔍 Cloudflare tested security-focused LLMs on its infrastructure and reports detailed findings from using Anthropic’s Mythos Preview as part of Project Glasswing. The model stood out for exploit chain construction and automated proof generation, producing runnable PoCs and iterating on failures. Its emergent guardrails proved inconsistent across runs and prompts, so Cloudflare built a tailored harness and additional safeguards to scale safely. The team also observed higher-quality, actionable findings compared with earlier frontier models, but noted increased noise from memory-unsafe languages and model bias.

🔒 The article argues that security must move beyond detection and focus on compressing the OODA loop—observe, orient, decide and act—so defenses can outrun attackers. It notes that detection improvements have reached diminishing returns while investigation and remediation remain time-bound bottlenecks. By embedding contextual investigation into systems and deploying agent-based remediation, teams can make faster, more consistent decisions. As AI-driven interactions accelerate threat timelines, continuous validation and automated response become essential.

🆕 AWS added FLUX.2-klein-base-4B and Qwen3-Embedding-0.6B to Amazon SageMaker JumpStart. FLUX.2 targets real-time image generation and multi-reference editing in a compact architecture that can run on consumer GPUs with about 13GB VRAM. Qwen3-Embedding delivers instruction-aware, multilingual text embeddings across 100+ languages for retrieval, RAG, and semantic search. Customers can deploy these models via SageMaker Studio or the SageMaker Python SDK.

🛡️ Microsoft Security explains how rising agentic autonomy reorients security from models to how agents are assembled, constrained, and governed inside applications. The post identifies amplified risks—agent hijacking, intent breaking, data leakage, supply chain compromise—and shows why the application layer is decisive because builders fully control permissions, tool access, and failure handling. It recommends concrete design patterns: agents as microservices, least permissions, deterministic human-in-the-loop, and distinct agent identity to limit blast radius and preserve auditability.

⚠️ AI hallucinations—confident but incorrect outputs—are increasingly driving risky decisions in critical infrastructure and cybersecurity operations, exploiting human trust in authoritative-sounding responses. A 2025 AA-Omniscience benchmark of 40 models found most systems were more likely to offer a confident wrong answer on difficult questions, underscoring that AI outputs must be treated as potential vulnerabilities until vetted. Effective controls include enforced human review before sensitive actions, treating training data as a security asset, strict least-privilege for AI systems, and prompt-engineering training to reduce ambiguous inputs.

🔍 Anthropic’s announcement that Claude Mythos Preview will not be released publicly underscores both genuine capability and strategic constraint. Independent testing and reproductions suggest similar performance from OpenAI’s GPT-5.5 and smaller community models, while Mythos’ cost and corporate incentives shape access. These generative systems dramatically improve automated vulnerability discovery, empowering both attackers and defenders. Mozilla’s use found 271 flaws, but many devices remain unpatchable, so organizations must adapt quickly.

🔒 Semperis finds that 93% of global organizations use or plan to use AI agents for security tasks such as password resets and VPN access, while 92% report AI on endpoints with SSH and encryption key access. The survey of 1,100 organizations warns of over‑permissioned and abandoned 'zombie' non‑human identities that increase hijack risk. Semperis recommends treating agents as NHIs, enforcing least‑privilege, and improving observability and recovery readiness.

🔒 Palo Alto Networks reports ongoing testing of frontier AI models, including Anthropic and OpenAI, finding they rapidly surface code vulnerabilities and potential exploit paths. In the May 'Patch Wednesday' advisories the majority of findings originated from these AI scans, prompting broad rescanning and remediation. The company warns of a narrow three-to-five-month window before AI-driven exploits spread and offers Unit 42 services to help organizations respond.

🛡️ Microsoft introduced MDASH (multi-model agentic scanning harness), a model-agnostic AI system in limited private preview designed to discover, validate, and prove exploitable defects in large codebases. The system orchestrates more than 100 specialized agents across frontier and distilled models in a structured pipeline that builds threat models, runs auditor and debater stages, groups equivalent findings, and proves vulnerabilities. Microsoft reports MDASH uncovered 16 issues fixed in this month’s Patch Tuesday, including two critical Windows networking and authentication flaws.

🔒 Anthropic's Mythos Preview, shared last month with a limited set of security partners, has demonstrated the ability to autonomously find zero-day vulnerabilities across major operating systems and browsers. Anthropic paired the release with Project Glasswing and $100 million in usage credits to help defenders, but reports of unauthorized access and denied requests from Chinese entities have already emerged. The development challenges the assumption of a durable US lead and has injected cybersecurity into high-level US–China summit talks, prompting urgent questions about access, regulation, and international cooperation.

🛡️ In a post-Mythos environment, AI-driven attacks can weaponize vulnerabilities within hours or minutes, outpacing traditional defensive cycles. Picus Security argues defenders must pair continuous Breach and Attack Simulation (BAS) with autonomous pentesting to validate controls and reveal genuine attack paths. Operational friction — the "spaghetti handoff" between tools and teams —, not tooling alone, is the main cause of delayed response, so validation must be automated end-to-end.

🔍 The UK’s AI Security Institute evaluated GPT-5.5’s ability to identify software security vulnerabilities and concluded it performs comparably to Claude Mythos, based on a series of red-team style tests and benchmark prompts. The assessment highlights that GPT-5.5 is generally available from OpenAI, making high-quality automated vulnerability detection more accessible to organizations and researchers. The Institute also analyzed a smaller, cheaper model which, when given additional prompting scaffolding and careful supervision, delivered similar detection performance. Overall, the study suggests parity among leading LLMs for initial vulnerability discovery, with differences largely hinging on prompt engineering and deployment context.

🔒 Microsoft researchers describe an AI-driven pipeline that translates attacker TTPs into realistic, structured security logs to accelerate detection engineering. The approach uses prompt engineering, collaborative agentic refinement, and data augmentation to generate semantically accurate telemetry (command lines, process ancestry, fields) without exposing sensitive customer data. Evaluation across multiple datasets shows agentic workflows and reasoning models notably improve recall and fidelity compared to prompt-only methods.

🔍 AI coding agents now operate across IDEs, terminals, and extension runtimes, so defenders must expand focus beyond source code to repository files, instruction and runtime settings, and third‑party extensions that shape agent behavior. VirusTotal Code Insight and agentic threat intelligence apply semantic analysis to detect malicious intent in syntactically valid artifacts and link findings to broader campaigns and supply‑chain risks. Examples—weaponized tasks.json, malicious Skill.md, redirected settings.json endpoints, and sabotaged extensions—illustrate how semantics can enable exfiltration, privilege escalation, and stealthy attacker control.

🔒 OpenAI has launched Daybreak, an initiative built on its frontier LLMs and the Codex assistant to help developers embed security throughout the software development lifecycle. Announced on May 12, Daybreak extends the Trusted Access for Cyber (TAC) program and includes GPT‑5.5, TAC-enabled GPT‑5.5, GPT‑5.5‑Cyber and a Codex Security research preview. The initiative supports code scanning, vulnerability triage, automated detection and response while pairing defensive capabilities with verification, proportional safeguards and accountability.

🔐 Agentic AI is already operating across enterprises, executing tasks and taking actions often without meaningful security involvement. Security teams must develop hands‑on fluency — build and test agents, understand integrations like the Model Context Protocol, and enforce scoped configurations — because policy alone won't close the gap. The piece distinguishes three agent classes (productivity, MCP‑connected vendor agents, and custom user agents) and emphasizes configuration, access scoping, and training such as SANS SEC545 to reduce exposure.

🔒 CISOs are shifting from a primarily technical control function to strategic business partners as AI reshapes risk, operations, and product delivery. Leaders such as Barry Hensley, Shaun Khalfan, and Jeff Trudeau stress publishing AI security frameworks, embedding security early in development, and aligning controls to business outcomes. They warn of AI-enabled threats — including advanced phishing, voice/video impersonation, and automated vulnerability discovery — and call for continuous controls, stronger identity and data governance, and near-real-time patching. Growing board engagement and changing reporting lines reflect the elevated role of security in enterprise strategy.

🛡️ OpenAI announced Daybreak, a cybersecurity initiative that combines GPT-5.5 family models with Codex Security to identify, test, and propose fixes for vulnerabilities before attackers exploit them. Daybreak builds editable threat models, runs isolated vulnerability tests, and suggests prioritized remediation and patch validation. Access is tightly controlled and available by request, and major vendors are integrating under Trusted Access for Cyber.

🔍 Google Threat Intelligence Group (GTIG) reports a rapid shift from nascent AI-enabled operations to industrial-scale use of generative models by threat actors. Based on Mandiant incident response, Gemini telemetry, and GTIG research, the report documents AI-assisted zero-day exploit development, autonomous malware like PROMPTSPY, and advanced obfuscation techniques. It highlights supply chain targeting of AI environments, anonymized premium LLM access, and specific interest from PRC- and DPRK-linked clusters. The report also outlines mitigations and defensive AI uses.

🔒 OTT Cybersecurity LLC — the team behind Lyrie.ai — announced acceptance into Anthropic’s Cyber Verification Program and the public release of the Agent Trust Protocol (ATP). ATP is an open cryptographic standard that enables real-time verification of an AI agent’s identity, authorized scope, attestation status, delegation, and revocation. The protocol is royalty-free, slated for IETF submission, and a reference implementation is published under an MIT license. Lyrie positions itself as the security layer for autonomous AI agents operating on the internet.