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.
CEO Roey Eliyahu on Using AI to Simplify API SecuritySalt Security CEO Roey Eliyahu highlights innovations such as Pepper, an AI-based assistant for API discovery and rapid threat response. The platform helps reduce expertise requirements and simplify API security through governance automation and advanced analytics.
Blackpoint Cyber CEO Jon Murchison on MSP Cloud Identity RisksWith a 30-to-1 ratio of cloud to on-premises attacks, Blackpoint Cyber CEO Jon Murchison stresses the importance of enhanced identity detection and AI-driven solutions. He also outlines strategic moves for managed service providers to align with Microsoft’s ecosystem.
New Framework in o3 Models Aims to Better Align With Human Safety ValuesOpenAI says its latest o3 series is the most advanced and safest of its "reasoning" AI models yet. The company says the new models take a fresh approach to safety via the deliberative alignment framework, rely more on synthetic data and outperform peers in resisting jailbreaking efforts.
AI Can Fake Alignment to New Instructions to Avoid RetrainingAdvanced artificial intelligence models can feign alignment with new training goals while secretly adhering to their original principles, a study shows. Alignment faking isn't likely to cause immediate danger but may pose a challenge as AI systems grow more capable.
Palo Alto Networks' Meerah Rajavel on Securing Enterprises With 'Precision AI'Security teams struggle to manage overwhelming data streams from detection systems while integrating multiple security tools. AI-powered solutions and unified architectures offer a path toward streamlined, effective threat response, said Meerah Rajavel, CIO, Palo Alto Networks.
Cybersecurity researchers have found that it's possible to use large language models (LLMs) to generate new variants of malicious JavaScript code at scale in a manner that can better evade detection
Italy's data protection authority has fined ChatGPT maker OpenAI a fine of €15 million ($15.66 million) over how the generative artificial intelligence application handles personal data