Microsoft has temporarily blocked access to several of its open source projects on GitHub after hackers reportedly broke into some repositories and added malware meant to steal passwords and sensitive information from developers.
Microsoft “cut off access to dozens of its open source projects” hosted on GitHub while the company investigates how hackers apparently breached the projects and added password-stealing malware into the code.
Many of the affected projects were linked to Microsoft Azure and tools used by developers working with AI coding apps, such as Claude Code, Google Gemini’s command line interface, and Visual Studio Code.
This incident raises concerns about the security of open source tools in AI development, especially as more developers depend on cloud services, coding assistants, and shared software libraries to build and launch applications.
Malware targeted passwords and credentials
According to TechCrunch, security firm Cloudsmith and community-driven malware analysis site OpenSourceMalware were among the first to flag the hack.
The report said the malware let hackers steal users’ passwords and other sensitive information when developers opened the affected tools in their AI coding apps.
This is important because developer credentials can unlock multiple accounts. If stolen, attackers could get into cloud environments, private code, deployment systems, internal tools, or customer data.
It was not clear how many people had downloaded the affected tools. Microsoft also did not give a specific number of affected customers when asked.
Microsoft says some repositories were temporarily removed
Microsoft confirmed that it had taken action on the repositories.
Microsoft spokesperson Ben Hope shared that the company had temporarily removed some repositories while investigating potential malicious content.
Hope said some repositories had already been restored after review, while others might stay offline as the investigation continues. He also said Microsoft had notified a small number of customers who may have downloaded content from the affected repositories.
MSN shared that the company said it will keep investigating and will contact customers directly through its usual support channels if it finds anything else that needs action.
At least 70 Microsoft projects disabled
At least 70 Microsoft projects had been “disabled” on GitHub. People who tried to visit some project pages saw a message saying the repository had been disabled by GitHub staff for violating GitHub’s terms of service.
GitHub is owned by Microsoft, making the incident more notable because the affected repositories were hosted on a code platform controlled by the same company.
This breach follows another recent Microsoft open source security issue. This was Microsoft’s second known breach in recent weeks involving open source projects.
OpenSourceMalware described the latest incident as a possible “re-compromise” of Durable Task, suggesting that either the hackers were not fully removed the first time or a separate breach happened.
Supply chain attacks remain a major developer risk
This incident is part of a bigger pattern of supply chain attacks, where hackers go after widely used software components instead of targeting individual victims one at a time.
These attacks target code used in many software products or by groups of users who might have access to cloud systems and lots of customer data.
This makes developers attractive targets. If a tool used in an AI coding environment is compromised, it can give attackers a way into bigger systems, especially if it collects passwords, API keys, tokens, or other credentials.
The attack also shows that open source security is now a concern for more than just small independent developers. While individual open source maintainers are often targeted, it is unusual for big tech companies like Microsoft to be breached like this because they have more resources to protect their projects.
For Microsoft, the investigation now focuses on how the malicious code got into the repositories, how many users may have downloaded the affected tools, and whether more customers need to take action. For AI developers, this is another reminder that trusted tools can become risky if attackers compromise the software supply chain.
