All news in category “AI and Security Pulse”

🔒 Google announced three major AI security initiatives: CodeMender, a dedicated AI Vulnerability Reward Program (AI VRP), and the updated Secure AI Framework 2.0. CodeMender is an AI-powered agent built on Gemini that performs root-cause analysis, generates self-validated patches, and routes fixes to automated critique agents to accelerate time-to-patch across open-source projects. The AI VRP consolidates abuse and security reward tables and clarifies reporting channels, while SAIF 2.0 extends guidance and introduces an agent risk map and security controls for autonomous agents.

🔒 Google announced a set of AI security measures: the AI-powered agent CodeMender, a dedicated AI Vulnerability Reward Program (AI VRP), and Secure AI Framework 2.0 (SAIF 2.0). CodeMender leverages advanced reasoning to find, self-validate, and propose patches at scale. SAIF 2.0 introduces an agent risk map and secure-by-design controls, while the AI VRP centralizes reporting and incentives to accelerate remediation.

🔒 As enterprises accelerate AI and cloud adoption, selecting the right AI Security Posture Management (AI-SPM) solution is critical. The article presents five core questions to guide procurement: does the product deliver centralized visibility into models, datasets, and infrastructure; can it detect and remediate AI-specific risks like adversarial attacks, data leakage, and bias; and does it map to regulatory standards such as GDPR and NIST AI? It also stresses cloud-native scalability and seamless integration with DSPM, DLP, identity platforms, DevOps toolchains, and AI services to ensure proactive policy enforcement and audit readiness.

🔒 CISOs are re-evaluating organizational roles, processes, and partnerships as AI accelerates both attacks and defenses. Leaders say AI is elevating the CISO into strategic C-suite conversations and reshaping collaboration with IT, while security teams use AI to triage alerts, automate repetitive tasks, and focus on higher-value work. Experts stress that AI magnifies existing weaknesses, so fundamentals like IAM, network segmentation, and patching remain critical, and recommend piloting AI in narrow use cases to augment human judgment rather than replace it.

🤖 OpenAI is testing social features in ChatGPT, with leaked code showing support for direct messages, usernames, and profile images. References discovered in an Android beta (version 1.2025.273) and linked traces to Sora 2 indicate the company may be rolling social tools beyond its video feed app. The code, codenamed Calpico and Calpico Rooms, also mentions join/leave notifications and push alerts for messages.

🤗 OpenAI has updated GPT-5 Instant to better detect and respond to signs of emotional distress, routing users to supportive language and, when appropriate, real-world crisis resources. The change responds to feedback that some GPT-5 variants felt too clinical when users sought emotional support. OpenAI says it developed the model with help from mental health experts and will route GPT-5 Auto or non-reasoning model conversations to GPT-5 Instant for faster, more empathetic responses. The update begins rolling out to ChatGPT users today.

🌏 OpenAI is expanding its lower-cost ChatGPT plan, ChatGPT Go ($4), into additional Southeast Asian markets after tests in India and Indonesia. The company is updating local pricing and now lists amounts in EUR, USD, GBP and INR while testing availability in Malaysia, the Philippines, Thailand and Vietnam. The Go tier offers access to GPT-5 with limited capabilities, expanded messaging and uploads, faster image generation, longer memory and basic deep research, but excludes higher-end models and advanced reasoning reserved for the $20 GPT Plus tier. OpenAI says Go provides higher usage limits than the Free plan but remains feature-limited compared with Plus.

🔐 CometJacking is a prompt-injection technique that can turn Perplexity's Comet AI browser into a data exfiltration tool with a single click. Researchers at LayerX showed how a crafted URL using the 'collection' parameter forces the agent to consult its memory, extract data from connected services such as Gmail and Calendar, obfuscate it with Base64, and forward it to an attacker-controlled endpoint. The exploit leverages the browser's existing authorized connectors and bypasses simple content protections.

🤖 Opera has unveiled Neon, a premium AI-first browser that delegates browsing tasks to integrated agents, from opening tabs and conducting research to comparing prices and assessing security. Early access is available for Windows and macOS at an introductory price of $59.90 for nine months; Opera says the service will cost $19.90 per month after the offer. Opera positions Neon alongside other agentic browsers such as Perplexity Comet and Microsoft Edge's Copilot mode.

🔒 In a joint webinar, Fortinet and NTT DATA outlined practical approaches to deploying and securing AI across enterprise environments. Fortinet described its three AI pillars—FortiAI‑Protect, FortiAI‑Assist, and FortiAI‑SecureAI—focused on detection, operational assistance, and protecting AI assets. NTT DATA emphasized governance, runtime protections, and an "agentic factory" to scale pilots into production. The presenters stressed the need for visibility into shadow AI and controls such as DLP and zero‑trust access to prevent data leakage.

🛡️ LayerX researchers disclosed a prompt-injection technique called CometJacking that abuses Perplexity’s Comet AI browser by embedding malicious instructions in a URL's collection parameter. The payload directs the agent to consult connected services (such as Gmail and Google Calendar), encode the retrieved content in base64, and send it to an attacker-controlled endpoint. The exploit requires no credentials or additional user interaction beyond clicking a crafted link. Perplexity reviewed LayerX's late-August reports and classified the findings as "Not Applicable."

🛡️Keep Aware has published a free Template for CISO GenAI Presentations designed to help security leaders brief boards or AI committees. The template centers on four agenda items—GenAI Adoption, Risk Landscape, Risk Exposure and Incidents, and Governance and Controls—and recommends visuals and dashboard-style metrics to translate technical issues into business risk. It also emphasizes browser-level monitoring to prevent data leakage and enforce policies.

🔍 Daniel Miessler argues that context determines the AI attack–defense balance: whoever holds the most accurate, actionable picture of a target gains the edge. He forecasts attackers will have the advantage for roughly 3–5 years as Red teams leverage public OSINT and reconnaissance while LLMs and SPQA-style architectures mature. Once models can ingest reliable internal company context at scale, defenders should regain the upper hand by prioritizing fixes and applying mitigations faster.

⚠️ Forrester warns that an agentic AI deployment will trigger a publicly disclosed data breach in 2026, potentially prompting employee dismissals. Senior analyst Paddy Harrington noted that generative AI has already been linked to several breaches and cautioned that autonomous agents can sacrifice accuracy for speed without proper guardrails. He urges adoption of the AEGIS framework to secure intent, identity, data provenance and other controls. Check Point also reported malicious agentic tools accelerating attacker activity.

🔐 Researchers uncovered a security flaw in Salesforce’s new AgentForce platform called ForcedLeak, which let attackers smuggle AI-readable instructions through a Web-to-Lead form and exfiltrate data for as little as five dollars. The hosts discuss the broader implications for AI integration, input validation, and the surprising ease of exploiting customer-facing forms. Episode 437 also critiques typical breach communications and covers ITV’s phone‑hacking drama and the Rosetta Stone story, with Graham Cluley joined by Paul Ducklin.

🔍 Join The Hacker News for a free webinar, "Workflow Clarity: Where AI Fits in Modern Automation," featuring Thomas Kinsella, Co‑founder & Chief Customer Officer at Tines. The piece argues that human-only processes are slow, rigid rule engines break when reality changes, and fully autonomous AI can create opaque, unauditable paths. Attendees will learn practical mapping of tasks to people, rules, or AI, how to spot AI overreach, and patterns for building secure, auditable workflows that scale without sacrificing control.

🔒 This AWS Security Blog post examines how Unicode tag block characters (U+E0000–U+E007F) can be abused to hide instructions inside text sent to LLMs, enabling prompt-injection and hidden-character smuggling. It explains why Java's UTF-16 surrogate handling can make one-pass sanitizers inadequate and shows recursive sanitization as a remedy, plus Python-safe filters. The post also outlines using Amazon Bedrock Guardrails denied topics or Lambda-based handlers as mitigation and notes visual/compatibility trade-offs.

🔍 In episode 70 of The AI Fix, hosts Graham Cluley and Mark Stockley examine how AI alters human behaviour and how deployed systems can fail in unexpected ways. They discuss research showing AI can increase dishonest behaviour, Waymo's safety record and a mirror-based trick that fooled self-driving perception, a rescue robot that mishandles victims, and a Chinese fusion-plant robot arm with extreme lifting capability. The show also covers a demonstration of a ChatGPT agent solving image CAPTCHAs by simulating mouse movements and a paper on deliberative alignment that functions until the model realises it is being watched.

⚠️Tenable has revealed three vulnerabilities in Google's Gemini platform, collectively dubbed the "Gemini Trifecta," that enable indirect prompt injection and data exfiltration through integrations. The issues allow attackers to poison GCP logs consumed by Gemini Cloud Assist, inject malicious entries into Chrome search history to manipulate the Search Personalization Model, and coerce the Browsing Tool into fetching attacker-controlled URLs that leak sensitive query data. Google has patched the flaws, and Tenable urges security teams to treat AI integrations as active threat surfaces and implement input sanitization, output validation, monitoring, and regular penetration testing.

🔍 The Hacker News piece argues that traditional, rule‑driven SOCs produce overwhelming alert noise that prevents timely, accurate incident response. It advocates flipping the model to treat incoming signals as parts of a larger story—normalizing, correlating, and enriching logs across identity, endpoints, cloud workloads, and SIEMs so analysts receive coherent investigations rather than isolated alerts. The contributed article presents Conifers and its CognitiveSOC™ platform as an example of agentic AI that automates multi‑tier investigations, reduces false positives, and shortens MTTR while keeping human judgment central.