Chinese AI industry leaders have acknowledged the need for China to develop its own hardware and software in order to catch up with the US in the field of generative AI. The focus is particularly on independently owned and controlled large language models (LLMs), which are the deep-learning AI algorithms used in chatbots like ChatGPT. Chinese companies such as iFlytek have launched their own chatbots but admit there is still a gap between Chinese and US LLMs in terms of product quality. Export restrictions on semiconductor chips, required for LLM processing, pose a major challenge to Chinese AI development. The US has prohibited chip sales to China due to national security concerns.

Chinese AI firms are advised to not only concentrate on LLM breakthroughs but also on developing new hardware and software. China and the US are in a fierce competition to dominate the AI field, and while China has made strides in facial recognition and autonomous driving, the introduction of OpenAI's text-to-video generator Sora has widened the gap once more. Chinese companies are attempting to close the gap with domestic rival chatbots like Baidu's Ernie Bot 4 and iFlytek Spark 3. 0. However, China needs to focus on the core and foundational technology of AI in order to compete. The biggest challenge to China's AI development lies in its risk-averse approach and reluctance to challenge the status quo.