This week’s Department of Know is hosted by Rich Stroffolinowith guests Janet HeinsCISO, ChenMedand TC NiedzialkowskiHead of IT & Security, Open door.
Missed the live show? Check it out on YouTube.
The Department of Know is live every Friday at 4:00 p.m. ET. Join us each week by registering for the open discussion at CISOSeries.com.
In this week’s cybersecurity news…
GitHub Drama
GitHub has been having a time of late. Checkmarx confirmed that data from its GitHub repository has been posted on the dark web following a March 23rd supply chain attack that compromised development tools and workflows. The breach involved tampered GitHub Actions and VS Code extensions distributing credential-stealing malware, with researchers linking subsequent leaks to groups like LAPSUS$ and activity attributed to TeamPCP.
Then GitHub disclosed a high-severity flaw that could let attackers with repository push access achieve remote code execution by injecting malicious metadata through unsanitized input. GitHub patched cloud instances with no evidence of exploitation, but many on-premise Enterprise Server deployments remained vulnerable.
And this comes as there is an increasing number of outages, leading to an increasing number of calls for changes, or outright leaving the platform. Mitchell Hashimoto summed this feeling up on a blog post, saying, “for the past month I’ve kept a journal where I put an ‘X’ next to every date where a GitHub outage has negatively impacted my ability to work. Almost every day has an X.”
(The Hacker News, Hashimoto, Dark Reading)
North Korean attacks use AI-inserted npm malware
Researchers at ReversingLabs are warning of malicious code in an npm package as a dependency to the project by Anthropic’s Claude Opus large language model (LLM). The package in question is “@validate-sdk/v2,” which is listed on npm as a utility software development kit (SDK) for hashing, validation, encoding/decoding, and secure random generation. “However, its real functionality is to plunder sensitive secrets from the compromised environment. The package, which shows signs of being vibe-coded using generative artificial intelligence (AI), was first uploaded to the repository in October 2025.” ReversingLabs has named this malware campaign PromptMink and has linked it to North Korean threat actor Famous Chollima.
HHS ponders government posture for protecting data centers
The question revolves around whether to designate data centers as a standalone critical infrastructure sector. Given that they are regularly targeted, a hearing was held on Wednesday to contemplate whether the federal government currently has the right setup for defending them. “Some industry witnesses and experts at the hearing of the House Homeland Security Subcommittee on Cybersecurity and Infrastructure Protection suggested that data centers be given their own standalone designation, especially in light of the boom in the building of such facilities across the country. This would follow a move already taken by the UK.
Huge thanks to our sponsor, Guardsquare
AI agent deletes production database… again
The founder of car rental SaaS platform PocketOS, Jer Crane, posted on X that an AI coding agent from Cursor deleted its “production database and all volume-level backups in a single API call to Railway,” the company’s infrastructure provider. The action took about 9 seconds. The Cursor agent was attempting to resolve a conflict by deleting a storage volume on Railway, using an API token found in a completed unrelated project. Railway also stored its backups on the same volume as the primary data source. PocketOS was able to restore from a full 3-month-old backup.
FIDO Alliance working on securing AI agent payments
The industry association said it’s working with Google and Mastercard on a pair of working groups to develop industry standards for validating and protecting payments made by AI agents. Google is contributing its Agent Payments Protocol to cryptographically verify that a user has authorized an agent. Mastercard will provide its Verifiable Intent framework, allowing users to authorize agents. The FIDO Alliance still needs to build out use cases for using both in real-world deployments, then work with merchants and payment providers on adoption and support.
(Wired)
Anthropic releases Claude Security Beta
We don’t cover many direct product releases on Cybersecurity Headlines, but I need your take on how to interpret this. Anthropic just released a beta of Claude Security to enterprise customers. This runs on Opus 4.7, allowing users to select a repository and scan for vulnerabilities. This will identify vulnerabilities, explain the findings, provide confidence-level information on severity, and generate instructions for a targeted patch. Anthropic also said “CrowdStrike, Microsoft Security, Palo Alto Networks, SentinelOne, TrendAI, and Wiz,” as well as the big consulting firms, are integrating Opus 4.7 into their platforms.
[ad_2]
Source link
