Anthropic has released a new AI model. It is weaker than Mythos, which caused concern in the Mythos markets

AI tool developer Anthropic has unveiled the Opus 4.7 model with improved software development capabilities. The company deliberately made it less powerful than its other new model Mythos: it remains in limited availability due to cybersecurity risks.

Anthropic announced on April 16 the launch of Claude Opus 4.7, an updated version of its most powerful widely available artificial intelligence model. The new version better handles complex programming tasks that previously required additional supervision and follows user instructions more accurately, the startup claims. The model has also gained improved high-resolution imaging capabilities, allowing it to analyze complex graphs and visual data.

At the same time, Opus 4.7 is inferior to Anthropic's other AI model called Mythos in terms of breadth of capabilities, especially in cybersecurity, Anthropic admitted. During the Opus 4.7 training process, the company intentionally reduced the model's abilities by experimenting with different constraints.

A week earlier, the startup reported that Mythos was capable of identifying and exploiting vulnerabilities in major operating systems and web browsers. As a result, access to it was restricted and made available only to selected corporate customers. Opus 4.7, for its part, is released with built-in protection mechanisms that automatically detect and block potentially dangerous cybersecurity-related requests.

Anthropic is stepping up competition with OpenAI, which created ChatGPT, in the AI segment by emphasizing programming and enterprise solutions, Bloomberg notes.

Since its founding in 2021, Anthropic has spent years carefully building its reputation as a company that is more committed than its competitors, including OpenAI, to the safe and responsible implementation of AI, CNBC writes. The heads of major U.S. banks, the Fed chairman and the U.S. Treasury secretary held a joint meeting last week on April 10 to address cybersecurity in the wake of the Mythos AI model, the network reported. "If it [Mythos] falls into the wrong hands, it could end really badly," European Central Bank Chair Christine Lagarde said.

Anthropic is increasingly popular with investors amid growing demand for the company's AI models, reflected in high rates of revenue growth. The startup has received several stock purchase offers from venture capitalists in the past two weeks, suggesting the company's business value is $800 billion or even higher, sources told Business Insider and Bloomberg.

This article was AI-translated and verified by a human editor