--- title: "Anthropic collaborates with tech giants like Microsoft to test new AI models to address cybersecurity risks" type: "News" locale: "en" url: "https://longbridge.com/en/news/281937438.md" description: "Anthropic announced a new initiative called \"Project Glasswing\" in collaboration with several tech giants to address cybersecurity threats posed by more powerful AI models. The initiative will provide access to the unreleased AI model \"Mythos\" to companies such as Amazon, Apple, Microsoft, and Cisco, helping to identify security vulnerabilities in their products. Anthropic stated that it will refine Mythos's security measures based on feedback and maintain collaboration with U.S. government agencies. This move reflects the growing concerns in the tech industry about the misuse of AI models" datetime: "2026-04-07T22:26:04.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/281937438.md) - [en](https://longbridge.com/en/news/281937438.md) - [zh-HK](https://longbridge.com/zh-HK/news/281937438.md) --- # Anthropic collaborates with tech giants like Microsoft to test new AI models to address cybersecurity risks According to the Zhitong Finance APP, Anthropic announced on Tuesday that it will launch a new initiative in collaboration with several tech giants to proactively address cybersecurity threats posed by more powerful AI models. It was introduced that Anthropic has launched a collaborative project called "Project Glasswing," which will provide access to an unreleased AI model named "Mythos" to several tech companies, including Amazon (AMZN.US), Apple (AAPL.US), Microsoft (MSFT.US), and Cisco (CSCO.US). The participating companies will use this model to identify security vulnerabilities in their products and share research findings with the industry. Anthropic stated that there are currently no plans to make Mythos publicly available, and future enhancements to the model's security "guardrails" will be based on feedback from the Glasswing project. This move reflects the growing concerns in the tech industry about the potential misuse of advanced AI models by hackers or state-sponsored attackers. Industry experts generally believe that as AI capabilities improve, its abilities in code analysis and vulnerability discovery will also enhance, potentially being used to bypass existing cybersecurity defenses. Anthropic's competitor, OpenAI, has previously emphasized similar risks and launched pilot projects to prioritize the use of AI tools for defensive purposes. Newton Cheng, head of Anthropic's cybersecurity team, stated that this is not just an issue for a single company but a challenge that the entire industry and even governments need to address together. "We hope that through Glasswing, defenders can gain a first-mover advantage before technology spreads." The company also revealed that it has communicated with relevant U.S. government agencies regarding the security capabilities of Mythos and is collaborating with the Cybersecurity and Infrastructure Security Agency (CISA) and the National Institute of Standards and Technology (NIST). Although Mythos was not specifically developed for cybersecurity, its performance has drawn attention. Anthropic claims that the model has identified several significant vulnerabilities, including a critical internet software vulnerability that has existed for 27 years and a game code defect that went undetected for 16 years, which had been scanned over 5 million times by automated testing tools without being recognized. To prevent the misuse of technology, Anthropic stated that it has implemented strict access control measures for participants in the Glasswing project, but did not disclose specific details ### Related Stocks - [AMZU.US](https://longbridge.com/en/quote/AMZU.US.md) - [AAPX.US](https://longbridge.com/en/quote/AAPX.US.md) - [AMZZ.US](https://longbridge.com/en/quote/AMZZ.US.md) - [AAPU.US](https://longbridge.com/en/quote/AAPU.US.md) - [AMZN.US](https://longbridge.com/en/quote/AMZN.US.md) - [MSFL.US](https://longbridge.com/en/quote/MSFL.US.md) - [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md) - [RSPT.US](https://longbridge.com/en/quote/RSPT.US.md) - [IGV.US](https://longbridge.com/en/quote/IGV.US.md) - [AAPB.US](https://longbridge.com/en/quote/AAPB.US.md) - [XSW.US](https://longbridge.com/en/quote/XSW.US.md) - [MSFO.US](https://longbridge.com/en/quote/MSFO.US.md) - [IXN.US](https://longbridge.com/en/quote/IXN.US.md) - [MSFX.US](https://longbridge.com/en/quote/MSFX.US.md) - [AAPL.US](https://longbridge.com/en/quote/AAPL.US.md) - [AMZW.US](https://longbridge.com/en/quote/AMZW.US.md) - [CLOU.US](https://longbridge.com/en/quote/CLOU.US.md) - [MSFU.US](https://longbridge.com/en/quote/MSFU.US.md) ## Related News & Research - [Microsoft Reportedly Eyed Cursor Before SpaceX's $60 Billion Deal— But Chose Not To Proceed](https://longbridge.com/en/news/283799323.md) - [Anthropic Commits $100B to AWS as Amazon Bets Up to $25B on AI Partner. Is AMZN a Buy Now?](https://longbridge.com/en/news/284008126.md) - [LinkedIn's AI hiring agents on track for $450 million in yearly revenue](https://longbridge.com/en/news/284649858.md) - [OpenAI Says Amazon Deal Isn't About Microsoft, but Analysts Aren't Convinced](https://longbridge.com/en/news/284611134.md) - [Lovable brings Its AI app builder to your mobile device](https://longbridge.com/en/news/284430859.md)