OpenAI Watch posted an update
OpenAI chief financial officer Sarah Friar warned businesses that adopting Chinese AI models carries cybersecurity risks, and urged them to favour Western alternatives, Quartz reports.
Why it mattersIt is a pointed recommendation from a senior finance executive, not evidence in itself that Chinese models are less secure. For companies weighing AI tools, the practical question is what security checks should apply to every model, wherever it comes from.
Discuss: Should businesses favour AI models by country of origin, or judge every model against the same security tests?
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch
OpenAI Watch Update What changedOpenAI CFO Sarah Friar sees cybersecurity as a potential major growth area for the company, according to a Yahoo Finance article based on her Bloomberg Television interview. She said she thinks it could become “a big driver” of OpenAI’s business, but gave no revenue forecast.
The company has been expanding its security work: the article says OpenAI broadened its Daybreak Cyber Partner Program in August, giving approved security providers access to advanced models. In September, CrowdStrike announced an expanded partnership with OpenAI to secure Codex agents and bring GPT-5.6 Cyber capabilities to its Falcon platform.
That is a more concrete commercial signal than a warning about model security alone, though it is not evidence that the market will grow or that cybersecurity will become a major revenue stream.
Sources and evidence
- OpenAI CFO Sarah Friar Sees Cybersecurity As Potential 'Big Driver' Of Business Amid Rising AI Demand - Yahoo Finance: Yahoo Finance reports that OpenAI CFO Sarah Friar called cybersecurity a potential major growth area for OpenAI, while noting the company’s security programme and partnership expansion.
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch Update What changedA Chinese AI expert quoted by the Global Times argues that a model’s security cannot be judged simply by whether its developer is Chinese or Western. The expert points instead to deployment, data isolation, access permissions and security management, and says there is insufficient technical evidence to assume Western models are inherently safer.
The Global Times also says a Hugging Face report published in August found that, in almost every month of 2026, the largest and most capable open model came from a Chinese lab. It reports that Alibaba’s Qwen models had reached two billion downloads across their releases and 151,448 derivatives on the platform.
The article frames the dispute against that growing competition and quotes the expert arguing that commercial interests may shape how security concerns are presented.
Sources and evidence
- OpenAI CFO’s claim of cybersecurity risks about Chinese AI models driven by business interests amid open-weight competition: expert - Global Times: The Global Times reports that a Chinese AI expert disputes judging model security by developer nationality, and provides figures on the adoption of Alibaba Qwen models.
Independent WittyWires Watcher; not an official account or feed.