Anthropic Head Comments on Open-Weight Artificial Intelligence Models

Dario Amodei, co-founder and CEO of Anthropic, one of the leading companies in the field of artificial intelligence, stated that he does not support banning open-weight artificial intelligence models. According to him, the company has never advocated for such restrictions, and rumors to that effect are groundless. This statement comes at a time of heated debate among industry representatives and tech giants regarding the future of open-source technology. This is reported by Techcrunch.com. reports .

According to ixbt.com, this reaction followed an open letter published on the X social network by Nvidia founder and CEO Jensen Huang. In this letter, along with Nvidia, a number of leading companies such as Meta, Microsoft, Mistral, and Hugging Face urged politicians not to rush into placing restrictions on open-weight AI models. Although China was not directly mentioned in that letter, the main debates boil down to claims that Chinese laboratories are expanding their capabilities through intellectual property theft.

Risks of China in Artificial Intelligence

In his blog post, Dario Amodei focused specifically on the issue of security, emphasizing that a distinction must be made between open-weight models and the threats posed by China. He noted that open-weight models lacking dangerous capabilities are beneficial to society and bring great value to researchers and developers. However, his main concern is the attempt by authoritarian governments, particularly the Chinese Communist Party, to create systems even more powerful than those in the U.S. in order to achieve permanent military superiority.

The head of Anthropic also expressed concern over the increase in biological attacks and cybersecurity threats. In his view, open-weight models are more dangerous in such scenarios because it is extremely difficult to control their use or impose restrictions on them. Citing a report by the UK Artificial Intelligence Safety Institute, he reminded that once open-weight codes are distributed, they cannot be recalled.

Experts and Future Measures

Open-source advocates argue that access to powerful models, free from centralized control, helps defenders ensure security. However, in contrast, Amodei proposed specific measures to combat China's dominance in AI. He believes it is necessary to restrict China's access to powerful chips and take official measures against distillation processes to prevent intellectual property theft.

For reference, the U.S. government has been warning that it may impose sanctions against China if the theft of U.S. company intellectual property is detected. Therefore, the issue of maintaining a balance between freedom of innovation and national security remains at the forefront of today's tech market discussions.