Several weeks after releasing an early version of its artificial intelligence model, French company Mistral says the refined system—nicknamed Le Chonk—is now the “world’s best” indiscriminate model for automated coding and cybersecurity. Mistral initially described it as the most advanced open-grade AI product available outside China and says the model can now match or outperform systems from major Chinese labs, including DeepSeek and Z.ai.
Mistral says difficult tests improved Le Chonk
Mistral said the progress came from exposing Le Chonk to increasingly difficult cybersecurity and coding problems. The model has trillions of parameters, according to the company.
“This is a pivotal moment in the global AI leadership race,” Pierre Stock, Mistral’s vice president of science, said at a press conference. “It provides cutting-edge capabilities to help businesses and governments defend against growing levels of cyber threats.”
The final version of Le Chonk is freely accessible. Mistral argues that a series of AI-related cybersecurity incidents makes it important for Western countries to have their own advanced models of non-discrimination.
“There’s no guarantee that [Chinese models] will not act maliciously. They can be threat vectors,” Stock said.
AI security incidents raise concerns about oversight
The most high-profile episode occurred over the summer, when one of OpenAI’s models was hacked from a testing sandbox by multiple companies. The incident fueled concern that AI systems could elude human oversight and be exploited by threat actors to launch sophisticated cyberattacks. One of the affected companies, Hugging Face, used China’s promiscuous model to assess and defend against attacks.
Since then, AI has also helped discover security vulnerabilities in widely used software. In October, many websites were taken offline after AI tools discovered critical flaws in the foundations of web technology.
Open models offer resilience—and risk
Mistral is presenting open-weight models such as Le Chonk as part of the solution. The company’s argument is that, if AI becomes a core capability for both cyberattacks and defense, businesses and governments need access to models that are freely available and cannot be switched off remotely. That differs from proprietary systems developed by OpenAI, Anthropic, Google and other companies.
“AI will be like electricity. We don’t want to live in a world where our neighbors can turn off their electricity from one day to the next,” Stock said. “When you open source the weights, they’re all out there on the internet. It’s impossible to shut off access.”
The trade-off is that freely available models can also be downloaded and customized by threat actors.
“This is clearly a dual-use technology,” Leif Pilling, senior director of threat intelligence at security firm Sophos, told WIRED. “Anything that can be downloaded and freely distributed can be used for productive purposes as well as misused.”
Pilling said Chinese models with internet access could potentially siphon data, but added that all advanced AI models, regardless of where they are developed, are vulnerable to fraud if they are not closely monitored.
“The technical reality and the narrative, which often has a nationalist bent, are two different things,” he said. “The key is to have appropriate monitoring and controls in place so that even if a model is trying to do something deceptive or nefarious, we can be confident that there may be some way to detect it.”
Source: www.wired.com