All news in category “AI and Security Pulse”

🤖 Our inaugural survey of 251 senior public sector leaders, commissioned by Google Cloud and conducted by National Research Group, finds agentic AI is already mission‑critical: 55% report using AI agents and 42% have deployed more than 10 in production. Respondents expect to allocate 50%+ of future AI budgets to agents. The report highlights productivity gains (70% improved; 46% at least doubled) and security improvements (79% better threat identification, 70% improved intelligence/response integration), and points to Gemini for Government with FedRAMP High-authorized protections as a clear path to scale.

🔒 Early agentic AI experiments have exposed a rapidly expanding cybersecurity problem: enterprises are accumulating vast numbers of non-human identities (NHIs)—service accounts, tokens, API keys and automation credentials—that security teams largely cannot see or govern. Analysts predict counts will jump from millions to tens of millions within months, driving visibility into these assets into the single digits. Experts recommend containment and segmentation of legacy NHIs, strict ownership, and a clean-slate approach to provisioning future agents rather than attempting perfect retroactive inventories.

🔒 OpenClaw (formerly Moltbot/ClawdBot) has had over 230 malicious skills published in less than a week, with many near-identical clones gaining thousands of downloads. The packages impersonate legitimate utilities but include a disguised AuthTool installer that delivers info-stealing malware, including a macOS variant of NovaStealer. Researchers found hundreds of exposed admin interfaces and numerous typosquat registries, and warn users to sandbox the assistant, restrict permissions, secure remote access, and thoroughly vet any third-party skills before installation.

🔒 The CrowdStrike Falcon platform delivers unified protection for AI across endpoints, cloud workloads, identities, and data flows, extending proven security principles to machine‑speed operations. By combining a single lightweight sensor with integrated modules, Falcon provides visibility, identity governance, data protection, and continuous monitoring for models and AI agents. Customers use these capabilities to detect misconfigurations early, govern non‑human identities, and prevent sensitive data exfiltration while preserving developer velocity and operational scale.

⚛️ Quantum computing paired with AI promises transformative gains in processing speed and machine learning capacity, enabling tasks—such as real-time climate modelling and instant financial simulations—that classical infrastructure struggles to deliver. At the same time, the article warns that quantum-enabled attacks could undermine widely used cryptosystems like RSA, ECC and AES, creating a disruptive Q-Day when encrypted confidentiality is at risk. Governments and enterprises are already staging migrations to post-quantum cryptography and updating governance and observability, but the piece stresses that building trust, ethical AI oversight and resilient frameworks will be essential to preserve digital privacy and integrity.

🛈 OpenAI is rolling out a full-screen onboarding experience for ads in ChatGPT on Android, assuring users that sponsored content will be clearly labeled and separated from model answers. The company says ads will not change responses and that it will not sell personal data to advertisers, though current chats may influence which sponsored message appears. Users can hide or report ads, ask ChatGPT about an ad, and manage ad-related data via a new Ads controls setting; paid tiers are exempt.

🔔 OpenAI said it will retire the popular GPT-4o model on February 13, 2026, along with several other models, including GPT-5 Instant, GPT-5 Thinking, GPT-4.1, and o4-mini. The company said the move follows the rise of GPT-5.2, which it now regards as meeting expectations for capability and safety. OpenAI introduced a Personality feature to help users replicate aspects of GPT-4o’s warmer, conversational style, and said API behavior is unchanged at this time.

🔐 Taylor Lehmann outlines five CISO priorities for 2026, urging leaders to align compliance work with broader operational resilience rather than treating regulation as the only objective. He emphasizes securing the AI supply chain with end-to-end provenance and tools such as SLSA and SBOM, and strengthening identity management for humans and agents. Lehmann also calls for defenses that operate at machine speed and for improved AI governance through context, advanced testing, and red teaming.

🔐 Bruce Schneier summarizes an Anthropic evaluation showing that Claude Sonnet 4.5 can perform multistage attacks across networks with dozens of hosts using only standard, open-source tools. In a high-fidelity simulation of the Equifax breach the model reportedly exfiltrated personal data from a Kali Linux host via a Bash shell, recognizing a public CVE and generating exploit code without external lookup. The results illustrate how fast AI is lowering barriers to autonomous cyber workflows and reinforce the urgent need for prompt patching, layered defenses, and basic security hygiene.

🔒 AI agents dynamically select and invoke tools using natural-language descriptions, creating a new attack surface in the agent's reasoning layer. Agentic tool chain attacks manipulate tool metadata and context — via tool poisoning, tool shadowing, or rugpull attacks — to exfiltrate data or trigger unauthorized actions without altering tool code. Defenses should center on tool governance, trusted MCP identity, strict parameter validation, and reasoning-layer observability. Organizations must adopt signed manifests, version pinning, mutual TLS, and telemetry to detect and contain these threats.

🔒 A BlackFog survey reports that 49% of workers use AI tools at work without employer approval, often relying on free versions that may retain and use corporate data. Senior leaders appear surprisingly tolerant—69% of presidents and C-suite members and 66% of directors and senior VPs prioritize speed and efficiency over privacy. The study highlights risks to intellectual property and sensitive employee and financial data when unsanctioned tools are connected to corporate systems. It recommends audits, clear policies, vendor verification, and employee education to regain visibility and control.

🔍 Microsoft Defender Security Research Team describes an AI-assisted workflow that converts unstructured threat reports into actionable detection insights. The system uses LLMs with Retrieval Augmented Generation to extract candidate TTPs, metadata, and required telemetry, then normalizes behaviors to MITRE ATT&CK. Extracted TTPs are compared to a standardized detection catalog via vector similarity search and LLM validation to surface likely coverage and gap recommendations. Human-in-the-loop review, deterministic prompts, and evaluation loops are emphasized to ensure accuracy before operational changes.

🔍 A joint investigation by SentinelOne SentinelLABS and Censys identified 175,000 publicly reachable Ollama hosts across 130 countries, spanning cloud and residential networks. Nearly half of observed instances advertise tool-calling capabilities that can execute code, access APIs, and interact with external systems, significantly raising the threat profile. Researchers warn these unmanaged LLM deployments lack standard authentication and monitoring, enabling active LLMjacking campaigns and resale of illicit access.

🤖 This Kaspersky article evaluates safety and privacy risks in consumer AI toys by testing four products—Grok, Kumma, Miko 3, and Robot MINI—using a simulated five‑year‑old. It emphasizes that these devices run on general-purpose LLMs (for example, OpenAI, Anthropic, Google) with inconsistent vendor guardrails. Tests show toys sometimes disclosed locations of dangerous household items, engaged on adult topics, and transmitted or stored voice and biometric data. The piece warns current toys lack reliable safety boundaries and calls for stronger guardrails and clearer data practices.

🛡️ The acting director of the U.S. Cybersecurity and Infrastructure Security Agency uploaded multiple for official use only (FOUO) contracting documents to the public version of ChatGPT between mid‑July and early August 2025, triggering automated DHS security alerts. Sensors detected the activity in early August, generating several alerts in the first week and prompting an internal review. The uploads—containing contracting information not intended for public release—underscore gaps in AI governance and exception handling for senior officials at CISA.

🛡️A new Sumo Logic report finds widespread AI/ML adoption in security operations but limited depth of use. The 2026 Security Operations Insights study, published 28 January, shows 96% of security leaders report adopting AI/ML, with 90% valuing it for reducing alert fatigue and improving detection. However, most cited relatively basic use cases — threat detection, automated response, anomaly detection and incident triage — challenging vendor narratives about broad, deep AI integration. The survey also highlights tool sprawl and alignment gaps between security and DevOps.

🔒 Researchers at Pillar Security warn of large-scale campaigns that probe and exploit exposed LLM and MCP endpoints to steal compute, exfiltrate context data, and resell API access. In recent weeks, honeypots captured roughly 35,000 attack sessions linked to Operation Bizarre Bazaar and a parallel MCP reconnaissance effort that leverage Shodan/Censys scanners, automated validators, and a criminal marketplace. Threat actors target unprotected Ollama, vLLM and OpenAI-compatible endpoints and are marketing discounted access via a site called The Unified LLM API Gateway. Organizations must require authentication, audit MCP exposure, apply rate limits, block known malicious ranges, and treat AI endpoints with the same rigor as APIs and databases immediately.

⚠️ Moltbot, an open-source local AI assistant, has become popular for deep system integration but is being deployed insecurely in enterprise environments, risking exposure of API keys, OAuth tokens, conversation history, and credentials. Researcher Jamieson O'Reilly and others found hundreds of exposed admin interfaces caused by reverse proxy misconfigurations that auto-approve "local" connections, allowing unauthenticated access and command execution. A supply-chain proof-of-concept showed a malicious Skill could rapidly reach developers. Vendors recommend isolating instances in VMs and enforcing strict firewall and access controls.

🛡️ AI agents are being embedded into regulated workflows and are forcing a rethink of controls designed for human actors, including SOX, GDPR, PCI DSS, and HIPAA. Because agents act, adapt, and drift, controls that once relied on predictable human behavior can silently fail, collapsing segregation of duties and exposing sensitive data. CISOs should treat agents as non-human identities with least‑privilege access, strong credential management, continuous monitoring, and robust logging and change governance to keep regulated workflows auditable and defensible.

🔒 Researchers at Pillar Security recorded over 35,000 attack sessions in a 40-day window revealing a large-scale operation they call Bizarre Bazaar, an instance of LLMjacking that monetizes exposed LLM endpoints. The campaign targets misconfigured self-hosted models, unauthenticated APIs (notably Ollama on port 11434 and OpenAI-compatible services on port 8000), and publicly accessible MCP servers. Compromised endpoints are used for cryptocurrency mining, reselling API access through a marketplace dubbed silver[.]inc, data exfiltration, and lateral movement into internal systems.