Articles about "malware"

Researchers have uncovered a malicious repository on Hugging Face that posed as an OpenAI release and distributed infostealer malware to Windows systems. The incident raises concerns about the security of AI model registries and the increasing sophistication of supply chain attacks targeting the AI ecosystem.

The rise of generative AI has empowered cybercriminals to automate and scale scams, including phishing emails and deepfake attacks. While these tools lower barriers for attackers, AI is also increasingly used for defense, as organizations race to keep pace with evolving threats. The issue is expected to intensify as AI technology advances and becomes more widely available.

Delve, a security compliance firm, recently assessed LiteLLM, an AI project that experienced a malware-related security incident. The situation highlights the growing importance of robust safety measures in AI development. The audit reflects increasing scrutiny of AI models and infrastructure following security breaches.

LiteLLM, an artificial intelligence project, suffered a malware breach after undergoing a security compliance assessment by Delve. The incident raises concerns about the effectiveness of compliance processes in detecting and preventing cybersecurity threats in AI development.

Microsoft reports that hackers are integrating artificial intelligence into every phase of their operations, reducing barriers to entry and enabling attacks at greater scale. The company highlights real-world examples, including North Korean groups using generative AI for fake job schemes and code generation. Microsoft warns organizations to treat AI-powered threats as insider risks and to harden systems accordingly.

Cybercriminals are leveraging fake Claude code downloads to spread malware among software developers. This campaign underscores mounting security concerns as generative AI platforms gain traction and become integral to development workflows. The incident heightens awareness about the risks associated with unauthorized AI tool downloads.

CrowdStrike’s 2026 Global Threat Report reveals an 89% year-over-year jump in AI-enabled cyberattacks, citing faster attack speeds and increasing targeting of AI systems themselves. State-backed groups, particularly from Russia, China, and North Korea, are using large language models (LLMs) and other AI tools to automate and accelerate intrusion tactics.

Google reported a decline in malicious Android apps targeting its Play Store in 2025, citing enhanced AI-driven security systems. The company blocked 1.75 million policy-violating apps and banned over 80,000 developer accounts as part of a broader push to safeguard users. Advanced generative AI models are now integrated into Google’s review process to detect threats more efficiently.

Google has reported that its artificial intelligence-driven systems effectively reduced the incidence of malware on the Google Play Store during 2025. The company attributes strengthened security measures to advances in machine learning and automated detection techniques. The announcement highlights AI's growing role in digital marketplace safety.

Google's Threat Intelligence Group has identified state-sponsored hackers from countries including Iran, North Korea, China, and Russia deploying artificial intelligence, such as large language models, in advanced cyberattacks. The report details how these actors use AI for reconnaissance, sophisticated phishing, malware development, and model extraction, elevating the cybersecurity threat landscape.