What is the risk of using Chinese open AI models like Kimi K3? - Financial Times

What the report says

The Financial Times, in an article headlined “What is the risk of using Chinese open AI models like Kimi K3?”, frames the debate over Chinese-developed open AI systems around cybersecurity and infrastructure resilience. Based on the available snippet, the article argues that the central concern is not simply that an open-source model was developed overseas, but that institutions lack sufficient coordination to defend critical systems against cyber attacks.

The limited public text does not identify a specific incident, regulatory decision or named organization beyond the reference to Kimi K3 and Chinese open AI models. It does, however, indicate a policy argument: focusing too narrowly on the foreign origin of AI tools may distract from broader weaknesses in how governments, infrastructure operators and technology providers prepare for and respond to cyber threats.

As background, open AI models can often be downloaded, modified and integrated into products or internal systems, which can make them attractive to developers and companies seeking lower costs or greater control. At the same time, their use can raise questions about software supply chains, model behavior, data governance and strategic dependence, especially when the technology comes from jurisdictions involved in wider geopolitical tensions.

The issue matters because AI tools are increasingly being tested or deployed in sensitive settings, including enterprise software and potentially critical infrastructure workflows. The FT’s reported emphasis suggests that risk management may require more than country-based restrictions; it may also depend on shared standards, security testing, incident planning and clearer responsibility across the organizations that operate essential services.