European firms face AI security and regulatory challenges

Chinese open‑weight AI models such as Kimi K3 and Alibaba's Qwen are gaining attention in Europe for their low token‑processing costs and open‑source nature. Experts argue that, while these models are not inherently more dangerous than other open‑source software, companies must still conduct security testing, bias checks and post‑training reviews before deployment. Concerns about potential back‑doors or data leakage have prompted U.S. tech giants to warn of competitive threats, but officials caution that the risk is largely theoretical.

In Poland, rapid AI adoption has created new privacy issues under GDPR. Up to 54 % of employees reportedly use AI tools without informing their employers, exposing personal and confidential data to external services. The forthcoming national AI Act, aligned with the EU AI Act, will add formal obligations for algorithm oversight, but businesses already face GDPR compliance challenges. A Polish tool designed to combat "shadow AI" offers a way to monitor and control unauthorized AI usage, helping firms set clear policies on data handling and responsibility.