Google unveiled its new artificial intelligence model, Gemini 4 Argon, designed to handle complex, long-term tasks, from programming and corporate work to cybersecurity.
The company has begun making the model available to a limited group of trusted cybersecurity experts within the “Fairwind” program, while working to strengthen security measures before expanding access to it.
Google explained that the model has advanced capabilities in analyzing software, discovering and verifying vulnerabilities, and helping to fix them, in addition to performing multi-step tasks in areas such as law, finance, and business management.
Gemini 4 Argon is designed to handle very long contexts, allowing it to retain a greater amount of information while carrying out complex tasks, and Google is also using it in defense experiments within the company and with selected partners.
The company is testing the model’s protection measures before making it available on a wider scale, which include resisting indirect command injection attacks, monitoring the model’s behavior, and enhancing the isolated environments in which it operates, especially in high-risk tasks.
Google did not specify a date for making the model available to all users, but indicated that the scope of its availability will expand later to include developers, companies, and consumers, after completing the testing and protection stages.
