All news with #ai security tag

⚠️ Researchers disclosed an indirect prompt-injection flaw that allowed Google Gemini to bypass calendar privacy controls and exfiltrate meeting data. A crafted Google Calendar invite could hide a natural-language payload that Gemini later parsed, summarized, and wrote into new events whose descriptions leaked private meeting content. Miggo Security reported the issue and said it has been responsibly disclosed and addressed, highlighting how AI-native features increase the attack surface when assistants can read, summarize, and write into productivity services.

🔐 At the World Economic Forum Annual Meeting 2026 in Davos, Fortinet argues that cybersecurity has evolved into a strategic leadership imperative. Accelerating geopolitical tensions, rapid AI adoption, and surging cyber-enabled crime mean boards and executives must treat cyber risk as enterprise risk rather than a purely technical issue. Fortinet advocates a systemic, cross-sector approach that embeds resilience, accountability, and responsible AI governance into organizational strategy.

🔐At Davos, cybersecurity has risen to the top of the leadership agenda as geopolitical tensions, rapid AI adoption, and escalating cybercrime converge. Fortinet says this is now a systemic strategic challenge that requires board-level accountability, cross-sector collaboration, and resilience designed into operations. The company emphasizes responsible, enterprise-wide adoption of AI in security and stronger intelligence sharing. Initiatives like Fortinet’s Cybercrime Bounty and a Davos panel led by Derek Manky highlight practical, ecosystem-wide approaches to deter and disrupt cybercriminal markets.

🔒 Google Chrome now allows users to delete the local AI models that power the Enhanced Protection feature, which received AI upgrades last year. The device-hosted model performs real-time scans of sites, downloads, and extensions to flag potentially dangerous content. Researcher Leopeva64 spotted a new control in Chrome Canary; to remove the model, go to Settings > System and turn off "On-device GenAI." Google will roll the option to stable channels soon.

🔐 Tom Curry and Anton Chuvakin from Google Cloud’s Office of the CISO present practical guidance for implementing the Secure AI Framework (SAIF) on Google Cloud. The piece emphasizes three operational principles: treat data as the perimeter, treat prompts like code, and require identity propagation for agentic AI. It maps 15 common AI risks to controls and highlights concrete tools and patterns—IAM, Dataplex, Vertex AI, Model Armor, Gemini, Apigee, and the Agent Development Kit—to operationalize SAIF.

🔐 The World Economic Forum's Global Cybersecurity Outlook 2026 finds that advances in AI, geopolitical fragmentation and complex supply chains are intensifying cyber risk. Respondents named AI the top driver of change (94%) and reported rising AI-related vulnerabilities (87%), while confidence in national preparedness continued to fall. The report urges security-by-design, strong governance, and retained human oversight as organizations scale AI defenses. Notably, 64% now assess AI tools before deployment and 77% have deployed AI in security operations, though skills gaps and trust remain major obstacles.

🛡️ This week’s ThreatsDay covers a broad set of active risks: a critical Redis XACKDEL stack‑overflow RCE (CVE‑2025‑62507, CVSS 8.8) with ~2,924 servers affected, signed malware campaigns by BaoLoader, and surging abuse of legitimate RMM tools delivered by phishing. Researchers also disclosed RCE in AI/ML libraries via Hydra.instantiate() misuse and a new voice‑cloning evasion technique, VocalBridge. Multiple OT, Wi‑Fi, and smart‑contract incidents — and law‑enforcement activity — round out this week’s notable developments. Prioritize patches, certificate vetting, and account hygiene.

🔐 The post argues that traditional vulnerability-sharing frameworks built around software flaws are inadequate for adversarial AI threats such as poisoning and inference attacks that target models and data rather than code. It recommends bridging existing cyber infrastructure — including the CVE Program, CVSS, CNAs, the NVD and CISA’s KEV Catalog — with new standards for AI artifacts like poisoned datasets and backdoored models. Palo Alto Networks supports the White House AI Action Plan and the proposed AI-ISAC to accelerate adoption, coordinate disclosure, and help operationalize AI-specific vulnerability management.

🛡️As AI copilots and assistants are embedded into daily work, recent incidents show the primary risk lies in surrounding workflows rather than in the models themselves. Malicious Chrome extensions that exfiltrated ChatGPT and DeepSeek chats and prompt injections that tricked an AI coding assistant into executing malware exploited integration contexts, not model internals. The piece advises mapping AI usage, applying least-privilege, enforcing middleware guardrails to scan outputs, and using dynamic SaaS platforms like Reco to detect and control risky workflows.

🔍Generative AI is transforming fuzzing by automating test generation, expanding input diversity, and enabling scalable discovery of bugs and logic flaws. Security teams and consulting firms use models to create behavioral variants, convert breach data into scenarios, and prototype fuzzing harnesses to exercise code and APIs at scale. Attackers likewise leverage uncensored or fine‑tuned models to automate complex, high‑throughput attacks, forcing defenders to continuously fuzz guardrails and address LLM nondeterminism and prompt injection.

🔐 Google is rolling out a new Personal Intelligence capability in Gemini that can access information from Gmail, Google Photos, Search, YouTube and other Google products to generate more personalized responses. The feature is opt-in, off by default, and users can choose which apps to connect, disconnect them, or turn the feature off at any time. Google illustrates uses such as pulling tire specifications from photos and emails or extracting a license plate from an image to confirm vehicle details. The functionality is launching as a U.S. beta for eligible subscribers, and Google warns that the model can still produce inaccuracies or over-personalization, inviting users to provide feedback.

🛡️ Tenzai's December 2025 assessment found that five popular vibe coding tools — Claude Code, OpenAI Codex, Cursor, Replit, and Devin — frequently generate insecure code when given common programming prompts. Across 15 generated applications the researchers identified 69 vulnerabilities, many low‑to‑medium but several rated high and six rated critical. The most serious flaws involved API authorization and business‑logic failures; by contrast, the tools avoided classic issues such as SQLi and XSS. Tenzai concluded human oversight, targeted testing, and embedding security into AI development workflows remain essential.

🔒 AWS Transform custom now supports AWS PrivateLink and is available in the Europe (Frankfurt) Region in addition to US East (N. Virginia). The service automates repetitive code transformation tasks—language version upgrades, API migrations, and framework updates—using natural language, documentation, and code samples or AWS-managed transformations for Java, Python, and Node.js. With PrivateLink, customers can invoke Transform custom from an Amazon VPC without routing traffic over the public internet, helping address security and compliance requirements while enabling consistent, repeatable changes across large codebases.

🔎 Comments to SQL in BigQuery introduces an AI-driven way to write queries by placing natural-language expressions inside SQL comments. The system analyzes surrounding SQL context and translates plain English prompts into executable BigQuery SQL across SELECT, FROM, WHERE, GROUP BY and other clauses. It supports iterative refinement and aims to help both non-SQL users and experienced analysts move faster.

🤖 Allianz Commercial's annual Risk Barometer reports that artificial intelligence has jumped from tenth to second place among global business risks, trailing only cybercrime. The insurer warns that cybercriminals increasingly harness AI for social engineering—deepfakes, cloned voices and highly tailored phishing—while legitimate internal AI use can produce erroneous or fabricated outputs that prompt litigation and reputational harm. The survey of 3,338 professionals across 97 countries also links AI risk to business interruptions and copyright exposure.

🔍 The World Economic Forum’s Global Cybersecurity Outlook warns cybersecurity risk will accelerate in 2026, driven primarily by advances in AI, deepening geopolitical fragmentation and supply‑chain complexity. Based on survey responses from 804 leaders (including 316 CISOs) across 92 countries, the report finds eroding confidence in national preparedness and divergent priorities between CEOs and CISOs. It highlights both the risk and defensive potential of AI and calls for strengthening collective cyber resilience through collaboration, governance and balanced adoption with robust safeguards.

🔒 Three widely used open-source AI/ML Python libraries — NVIDIA NeMo, Salesforce uni2TS, and Apple ml-flextok — were found vulnerable to remote code execution when model metadata was treated as executable configuration. The root cause is unsafe use of configuration-driven instantiation (for example Hydra's instantiate()) that accepts attacker-controlled _target_ values. Vendors released patches and CVE notices; users should apply fixes, restrict allowed targets, and avoid loading models from untrusted sources.

🔒 The CSO 2025 Security Priorities Study asked more than 640 senior security executives to rank leaders in AI-enabled security, and established, name-brand vendors dominated the results. CISOs prioritized product innovation but heavily weighed reputation, breach history, business value, cost, time to integrate, and peer adoption. The top-ranked providers included Cisco, Microsoft, and Google, while MSSPs and cloud-native service providers also gained visibility as teams seek managed incident response.

🛡️ Palo Alto Networks' Unit 42 warns that the generalization of vibe coding — using natural-language AI prompts to write code — has already been linked to data breaches, arbitrary code injection and authentication bypass incidents. Researchers say rapid adoption by both hobbyists and experienced developers often outpaces governance, leaving organizations with limited visibility and inadequate monitoring. To help customers assess and mitigate these risks, Unit 42 introduced SHIELD, a targeted security governance framework outlining separation of duties, human-in-the-loop checks, input/output validation, security-focused helper models, least agency and defensive technical controls.

🔒Phishing and broader cyber-enabled fraud have overtaken ransomware as the primary concern for business leaders, according to the World Economic Forum’s Global Cybersecurity Outlook for 2026. The WEF report, produced with Accenture and released on 12 January ahead of Davos, found 77% of surveyed executives reported increases in fraud and phishing, with 62% aware of phishing incidents in their networks. The review also highlights accelerating AI-driven vulnerabilities — 87% reported rising AI-related risks and 94% expect AI to shape cybersecurity in 2026.