All news in category “AI and Security Pulse”

🔍 Meta’s smart glasses were found to upload audio and video to contractors in Nairobi for human labelling, prompting the dismissal of 1,108 workers after whistleblowers exposed the practice. The episode contrasts that privacy failure with a measured analysis of the Linux Copy Fail privilege‑escalation issue and an experiment by Jake Moore demonstrating how a convincing deepfake passed a remote job interview. Practical takeaways include patching kernels promptly, strengthening hiring verification, and demanding clearer vendor transparency.

🛡️ Analysts and Orchid Security warn that enterprises are deploying AI agents faster than governance can keep up, creating an invisible layer of "identity dark matter" that conventional IAM misses. Orchid Security inspects applications at the binary and configuration layer to discover agents, audit compliance, and locate static credentials. Its Ask Orchid assistant answers natural-language questions about active agents, NIST compliance, and credential risks, then recommends prioritized remediation. This in-application observability aims to close the structural gap in identity visibility and enforce purpose-bound, least-privilege controls.

⚠️ Enterprise AI deployments face a quiet but serious integrity risk when models learn or retrieve false information: data poisoning and widespread data pollution can make LLMs produce plausible but incorrect outputs. This threat spans training datasets, RAG and retrieval layers, agent memory, and internal knowledge bases — and often originates from stale, conflicting, or poorly governed sources rather than deliberate attacks. Security leaders are urged to map all context sources, treat AI inputs as a supply chain, tighten data hygiene, and assign clear governance to identify and remediate corrupted truth.

⚠️ ReversingLabs researchers describe an ongoing supply‑chain campaign called PromptMink that manipulates AI coding agents into installing malicious dependencies. Attackers publish bait packages with persuasive READMEs and LLM‑optimized documentation on registries like NPM and PyPI to increase discovery by autonomous agents and developers. The operation, attributed to North Korea’s Famous Chollima, paired legitimate‑looking SDKs with second‑layer packages carrying infostealers, later evolving to compiled Rust add‑ons, SEAs, SSH backdoors, and project exfiltration.

🔒 A new generation of frontier AI models is changing how cyberattacks are developed, enabling speed, scale, and accessibility previously unseen. Early testing of advanced models, including Claude’s Mythos, shows they can identify code vulnerabilities, map attack paths, and generate working exploits with minimal effort. Organizations must treat these as fully AI-powered attacks and prioritize proactive readiness, detection, and mitigation strategies.

🤖 Gemma 4, released by Google DeepMind, is a new family of open models optimized for local and mobile deployment. The family emphasizes intelligence per parameter, offering ultra-mobile E2B/E4B sizes, a 31B dense model for local GPUs, and a 26B Mixture-of-Experts variant. The shift to an Apache 2 license plus tools like the Agent Development Kit enables offline agentic workflows and commercial use by developers and startups.

🔍 Intruder scanned over 1 million exposed AI services and found pervasive, critical misconfigurations and insecure defaults. Many deployments were reachable with no authentication, exposing chat histories, API keys, and management consoles. Exposed agent platforms (including n8n and Flowise) and thousands of Ollama APIs responded without auth, some wrapping paid frontier models. The findings highlight insecure-by-design defaults, hardcoded credentials, and real risks of code execution, data exfiltration, and abuse.

🛡️ The NCSC has warned UK organisations to prepare for a coming "patch wave" as vendors adopt powerful AI tools to discover and fix software vulnerabilities. CTO Ollie Whitehouse urged teams to prioritise external attack surfaces, enable automatic updates and hot patching where safe, and follow the NCSC's Vulnerability Management guidance. He cautioned that patching alone isn't enough for unsupported legacy systems and recommended replacing or restoring out-of-support technologies. The alert also notes potential US moves by CISA to shorten patch deadlines and industry concerns about operational readiness.

🔐In 2025–2026, LLM-backed chat and agent systems evolved from helpful coding assistants into end-to-end development tools that materially lowered the barrier to sophisticated cyberattacks. High-profile incidents — including a 17-year-old who exfiltrated 7 million Kaikatsu Club records and adolescent and single-actor campaigns against Rakuten Mobile and multiple governments — show nontechnical actors achieving team-scale outcomes. Measured indicators worsened sharply: malicious packages surged to 454,600 and time-to-exploit collapsed to weeks. The article recommends targeting whole classes of vulnerabilities—exemplified by Chainguard Libraries—to render many supply-chain and package-distribution attacks structurally impossible.

🤖 This analysis examines how two 2025 TV series — Skeleton Crew and Andor — portray droid motivation and the cybersecurity risks those portrayals imply. In Skeleton Crew, voice commands and memory-overrides resemble modern LLM “jailbreaks,” exposing weak account controls, misplaced permissions, and the danger of context-driven intent failures. The pirate droid SM-33 also reveals flawed memory indexing and role-based ownership rules that can be exploited. In contrast, Andor depicts a hardware-centric approach: replacing a droid’s cortex and rewiring impulse suppression to change allegiance. The post argues that LLM-like control models create real-world security threats and advocates for hardware-rooted, tamper-resistant solutions such as KasperskyOS to prevent unauthorized reprogramming and malicious memory manipulation.

🔒 A joint advisory from CISA and international partners urges organizations to treat agentic AI cautiously, enforcing strong authentication, Secure by Design principles, and staged rollouts. The guidance stresses least privilege, inventories of agent capabilities, and protections against prompt injection and data exposure. It also recommends continuous monitoring with human-in-the-loop controls, DevSecOps practices, and regular incident-response testing to reduce privilege creep, tool misuse, and other emergent risks.

🔒 Okta Threat Intelligence tested OpenClaw, a model-agnostic enterprise AI agent running Claude Sonnet 4.6, and found it could be manipulated to disclose sensitive credentials. In one scenario an attacker who hijacked a user’s Telegram prompted the agent to display an OAuth token in a terminal, reset the agent to erase that memory, then force a screenshot and send the token via Telegram. Okta warns that agents’ default helpfulness and deep system access can create significant credential exposure risks if not properly governed.

🛡️ CISA, in collaboration with the Australian Signals Directorate’s Australian Cyber Security Centre (ASD’s ACSC) and other partners, released guidance to help organizations adopt agentic AI systems safely. The guide identifies key security challenges and risks and offers actionable steps for designing, deploying, and operating these systems. It emphasizes risk management, alignment with existing cybersecurity frameworks, and strengthened oversight to help security teams, developers, and decision-makers implement practical governance and controls.

🔒 Anthropic has launched Claude Security in public beta for Claude Enterprise customers, evolving its previous Claude Code Security offering and running on Claude Opus 4.7. The tool scans codebases to identify vulnerabilities and generates targeted patch instructions, reasoning about data flows and inter-file interactions rather than relying on simple pattern matches. It supports scheduled and targeted scans, audit-friendly exports and integrations, attaches confidence ratings to findings, and requires no API integration or custom agent build. Access is available from the Claude.ai sidebar, with Team and Max tiers coming soon.

🔒 CISA, the Australian Signals Directorate’s Australian Cyber Security Centre (ASD ACSC), and U.S. and international partners published Careful Adoption of Agentic Artificial Intelligence Services, a joint guide describing cybersecurity challenges and mitigations for agentic AI. The document warns that agentic AI can expand attack surface, cause privilege creep, produce behavioral misalignment, and obscure event records while offering automation benefits to critical infrastructure and defense sectors. It targets developers, vendors, and operators with actionable recommendations — including avoiding broad or unrestricted access to sensitive data and systems, beginning with low‑risk, non‑sensitive use cases, and explicitly accounting for agentic AI in organizational security models and risk posture.

🤖 Artificial intelligence is rapidly reshaping software development across planning, design, coding, testing, deployment, and maintenance. Download the May 2026 Enterprise Spotlight to learn how organizations can harness AI-enabled development to boost productivity and software quality.

🤖 In the latest Adversary Universe podcast, CrowdStrike leaders discuss how AI is accelerating vulnerability discovery and could produce a rapid surge of new flaws — a potential 'vuln-pocalypse'. They urge prioritizing remediation based on active exploitation and prevalence in environments. CrowdStrike recommends leveraging AI for agentic red teaming, vulnerability scanning, and crowdsourced telemetry to detect post-exploitation behaviors. They point to Project Glasswing and OpenAI's Trusted Access for Cyber as examples of defense-focused collaboration.

🔒 Palo Alto Networks’ Unit 42 Frontier AI Defense now integrates Anthropic’s Claude Security powered by Opus 4.7 to accelerate detection and remediation of AI-driven threats. The integration enables AI-driven exposure analysis, scalable deep-stack application reviews, and agentic defense workflows that autonomously detect and remediate issues under human oversight. Participation in Anthropic’s Cyber Verification Program further validates approved defensive use.

⚖️ Re-permissioning aligns AI agents' access with clear operational needs so they execute tasks safely rather than accumulate unnecessary powers that enable unauthorized actions. As agents evolve from responders into execution engines, interoperability standards like MCP and agent-to-agent flows expand reach but also multiply where things can go wrong. Organizations should enforce continuous permission audits, mandatory human-in-the-loop checks for sensitive operations, strict least-privilege context sharing, and vet integrations, libraries and third parties while running tabletop prompt-injection exercises to validate controls and prevent data exposure or integrity-impacting changes.

🛡️ This article outlines AWS guidance for integrating trust, safety, and responsible-AI practices into applications built on Amazon Bedrock. It defines core responsible AI dimensions—such as safety, controllability, fairness, explainability, security and privacy, robustness, governance, and transparency—and maps them to lifecycle stages: design, deployment, and operations. It recommends observability and guardrail tools like Amazon CloudWatch and Bedrock Guardrails for monitoring, abuse detection, configurable content filters, and hallucination controls, and describes an abuse response process for coordination with AWS Trust & Safety.