Explore the intersection of AI and cybersecurity. Stay informed on AI-driven security trends, tools, and threats in the ever-evolving digital landscape.
Search across headline titles and summaries.
Background for this topic.
Artificial intelligence (AI) describes computer systems that perform tasks such as recognizing patterns, making predictions, understanding language, or generating content. In security reporting, the term commonly includes machine-learning models used for detection and analysis, as well as generative AI applications that produce text, code, images, or other outputs.
AI can help analyze security telemetry, prioritize vulnerabilities, and support investigations, but its outputs can be wrong or manipulated. Important attack surfaces include prompt injection that steers an application into unintended actions, sensitive data being exposed through prompts or model outputs, and excessive permissions granted to AI systems that use external tools. Models can also be degraded by poisoned training data or evaded with carefully crafted inputs. Practitioners should protect training and operational data, limit model access and tool permissions, test for adversarial behavior, and require appropriate human validation before high-impact decisions.
Passing the buck, and the blame, down the road shows lack of AI companies' maturity OPINION AI vendors: "You need to use AI to fight AI threats (and do everything else in your corporate IT environment)." Also AI vendors: "That's not a security flaw; it's working as intended."…
In 2024, compromised service accounts and forgotten API keys were behind 68% of cloud breaches. Not phishing. Not weak passwords. Unmanaged non-human identities that nobody was watching
Officials Warned New Models Could Accelerate Cyber Risks Faster Than RulesGlobal finance officials meeting in Washington warned that advanced artificial intelligence models could expose structural weaknesses across banking and payment systems, speeding vulnerability discovery and cyber exploitation faster than regulators can build guardrails.
Optimizing Value and Utility Hinges on AI Scaffolding, Says Aisle's Ondrej VlcekWhile the world is in "awe" of how Mythos can find vulnerabilities and chain together exploits, the next step is to identify how to build the best cybersecurity pipelines and scaffolding to get maximum value from all AI models used inside an organization, said Aisle CEO Ondrej Vlcek.
Also: NY State Regs Test Resilience vs Compliance, OT Security Nears Breaking PointIn this week's panel, four ISMG editors explore the industry's response to Anthropic's Mythos AI breakthrough, whether tighter New York state cybersecurity rules are driving real resilience or simply compliance, and why operational technology security is fast becoming a critical frontline concern.
AI's danger isn't that it's creating new bugs, it's that it's amplifying old ones.
AI models are making rapid gains in vulnerability research and exploit development, raising new cybersecurity risks, a Forescout study finds
Equifax CTO Jamil Farshchi on Cybersecurity's Response to Flood of VulnerabilitiesCybersecurity organizations must adapt to machine-speed threats in the age of Anthropic's Claude Mythos, a new AI model that can uncover vulnerabilities and lead to a flood of repaid exploits. Equifax CTO Jamil Farshchi says security programs must be built for scale, automation and quick response.
Bank of America, Citi and Goldman Anchor Partner Cohort for OpenAI's GPT-5.4-CyberOpenAI's Trusted Access for Cyber program prioritizes financial institutions to drive adoption of GPT-5.4-Cyber in regulated environments, highlighting a split with Anthropic’s developer-centric, tech-heavy partnerships and raising questions about partnership value and data-sharing models.
Ising Model Family Targets Calibration, Error Correction at ScaleNvidia released what it calls the world's first family of open AI models built to reduce errors in quantum computers in a bid to tackle problems blocking the technology's path to practical use. Nvidia is focusing on software, AI models and tools that work alongside quantum hardware.
Google says it is increasingly using its Gemini AI models to detect and block harmful ads on its advertising platforms, as scammers and threat actors continue to evolve their tactics to evade detection. [...]
A new cybercrime platform called ATHR can harvest credentials via fully automated voice phishing attacks that use both human operators and AI agents for the social engineering phase. [...]
AI-powered SOC tools promise automation, but most only speed up triage instead of reducing real workload. Tines shows how real gains come from end-to-end workflows that execute actions across systems, not just summarize alerts. [...]
Forged metadata made AI reviewer treat hostile changes as though they came from known maintainer Security boffins say Anthropic's Claude can be tricked into approving malicious code with just two Git commands by spoofing a trusted developer's identity.…
Three incidents. No actual breaches. Full-scale crisis response. AI hallucinations are creating a new threat vector that most organizations have yet to prepare for. The post Ghost breaches: How AI-mediated narratives have become a new threat vector appeared first on CyberScoop.
Playbook Aims to Help Healthcare, Public Sector Manage AI Vendor Security GapsThe Health Sector Coordinating Council released guidance to help the healthcare and public health sector better manage the explosion of third-party AI vendor cyber risk concerns they face, especially as the technology is embedded in all sorts of products.
Agencies Prioritizing Tracking Use Over Enforcing Immediate CutoffsFederal agency staffers tell ISMG they are still using Anthropic's AI tools weeks after U.S. President Donald Trump ordered an immediate halt, as officials prioritize mapping dependencies and evaluating alternatives over enforcing a rapid shutdown.
Felicis-Led Series A Backs Telemetry Correlation Across Cloud, Identity, EndpointsArtemis, a New York startup led by former Amazon GuardDuty product leader Shachar Hirshberg, emerged from stealth with $70 million to build an AI-driven SIEM alternative that correlates telemetry across enterprise environments, tailors detections and speeds investigations.