Six METR employees can examine internal tools and safety practices, with publication rights designed to let them report risks without Anthropic’s approval. Anthropic has given six employees of AI evaluation nonprofit METR access to internal systems and tools comparable to those used by its own risk assessment teams, in a move designed to give independent […]
Continue Reading
IBM and Red Hat’s Lightwell service pairs a $712 million savings model with embargoed fixes for paying customers. Security teams still have little public data to verify the claims More than 400 previously unknown vulnerabilities in widely used Java libraries have been fixed through a new IBM and Red Hat remediation service. Only one of […]
Continue Reading
Cisco Talos and Recorded Future identified four malware delivery methods and exposed infrastructure linked to the campaign, which researchers assess may have a Kazakhstan connection. A threat group targeting government organizations across Central Asia and the Commonwealth of Independent States (CIS) has used four different methods to deliver malware, including a WinRAR vulnerability exploit, while […]
Continue Reading
The AI model can identify and help fix complex software vulnerabilities, but Google is restricting access to its less-guarded cybersecurity capabilities through its Fairwind Program Google is expanding access to Gemini 4 Argon, its latest AI model for cybersecurity defense, to hundreds of additional vetted partners as the company seeks to help security teams identify […]
Continue Reading
The ecosystem combines phishing kits with an underground marketplace selling compromised cloud environments, enabling attacks against organizations in more than 150 countries. Microsoft has uncovered a cybercrime supply chain that combines phishing-as-a-service kits with an underground marketplace selling compromised cloud environments, enabling criminals to send more than one billion phishing emails and SMS messages to […]
Continue Reading
Aikido flags self-propagating malware in trusted packages, with execution shifting from install-time to runtime behavior A compromised npm package linked to Memtensor is being used to distribute a self-propagating supply chain worm capable of spreading across repositories and package registries, according to security firm Aikido. The malware, also identified in a related PyPI package, executes […]
Continue Reading
The company claims new model reduces operating costs by up to 40% while improving performance and resistance to prompt injection Anthropic has introduced Claude Opus 5.5, a new flagship AI model that the company said delivers improved performance while reducing compute costs and strengthening alignment safeguards. The release marks the first model in Anthropic’s Claude […]
Continue Reading
FBI, UK NCSC and Dutch AIVD warn Chosen Brick spyware survives reboots and harvests contacts, messages and screen data from compromised systems Iranian state-linked cyber actors are using popular messaging apps to trick targets into installing surveillance malware on their Windows machines, three Western government agencies warned in a joint advisory published Tuesday. The FBI, […]
Continue Reading
Broadcom also joins SE Labs’ new PIVOT programme, as participation in the US-backed MITRE Engenuity evaluations falls from 30 vendors to 11 A wave of major cybersecurity vendors is shifting away from the US Department of Defense-backed MITRE Engenuity ATT&CK Evaluations toward a new independent testing programme run by British company SE Labs, according to […]
Continue Reading