The phenomenon Deepseek It emerged just over a month ago and since then this Chinese AI Development Company He has not stopped filling covers, largely due to his potential to leave behind the rest of the competition in this industry. That capacity It could especially affect Silicon Valley and, particularly, Openaithe company that has managed to popularize generative artificial intelligence thanks to Chatgpt.
Sam Altman, CEO of OpenAIHe has publicly recognized the advances of Deepseek and his artificial intelligence model R1. In an X publication (previously Twitter), described this technology as “impressive”emphasize its ability to offer high performance at a reduced cost.
However, Openai has also officially expressed concerns about the rapid evolution of Deepseek and even accused him of using his models as a training method.
Now, In a letter signed by the Vice President of Global Affairs, Chris Lehane, and addressed to the United States Government, Altman’s company warns about Depseek’s R1 model, stating that it could represent security risks. OpenAi says openly that, due to pressures of the Chinese government, it is possible to be manipulated.
Openai has sent this statement to the White House and Technology Office In response to an information request on the development of an action plan for artificial intelligence. The company argues that the US must maintain its leadership in front of China’s competition, specifically the Deepseek R1 model, which they consider a strategic challenge promoted by the Chinese Communist Party (PCC).
OpenAi’s accusations Deepseek: this is “subsidized and controlled” by the Chinese state
“In the advance of democratic AI, The United States competes with a PCC determined to become the world leader for 2030. Therefore, the recent launch of Depseek’s R1 model is so remarkable, not because of its capacities (the reasoning capacity of the R1, although impressive, is, in the best case, along with that of several American models), but as an indicator of the state of this competition, ”says Openai in his letter.
In this statement, the Altman company makes such categorical statements as There is the possibility that “Deepseek is forced by the PCC to manipulate its models to cause damage” and that this is “subsidized and controlled” by the Chinese state. Therefore, they add, “the cost for their users is their privacy and security.”
OpenAi’s proposal to the United States: Frena the Chinese AI
Depseek is not the only one of Chinese origin that could make things difficult to OpenAi and, in general, to the artificial intelligence industry in the United States. Manus is the last feeling in the sector and, only days after its launch, it already has thousands of applications to prove it.
For this reason, the Altman company proposes a series of strategies for the US to continue leading the AI and do not lose competitive advantage.
Among them, urgent create a regulatory framework that “foster innovation” and does not impose “excessive regulations”. Openai warns that multiple United States state laws on AI are fragmenting the market, which could benefit China.
In addition, advocates export control to protect “democratic.”. The concept of ‘democratic’ (democratic ai) appears several times in the letter. Openai uses it to refer to an artificial intelligence approach based on democratic values and opposite to what they consider an authoritarian model promoted by the Chinese Communist Party.
For that export control, They suggest differentiating countries into three levels:
- Tier I: Ee. Uu. Uu. That respect democratic principles in AI.
- Tier II: Countries with doubtful history of technology exports to China, but can improve with support from the United States.
- Tier III: China and other allied countries of the PCC, which should be completely excluded from the access to American advanced.
They also mention that the use of Chinese chips and AI models with security vulnerabilities in critical infrastructures of the United States should be prohibited.
Altman’s company defends the use of protected data by Copyrightthat is, reforms in the field of copyright to protect data access.
Finally, Openai insists that the US needs massive investments in data centers, electrical transmission and connectivity to support the needs that artificial intelligence will lead to and seeks to occur what calls a ‘Pact for AI’ with United States allies to build a global technological infrastructure aligned with democratic values. It suggests that there are tax incentives, credits and regulatory exemptions to accelerate infrastructure construction.
Sign up to us Newsletter And receive the latest news about technology in your mail.
#OpenAi #charges #China #Depseek #cost #users #privacy #security
