The Biden Administration relaunches the challenge on the development of Artificial Intelligence and pushes seven large companies competing on AI to sign security guarantees on the development of the technology. The companies are Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI.
Biden announced the agreement in a short speech from the Roosevelt Room in which he also reiterated that this is only a first step and that his administration is trying to finalize an executive order on the AI issue. In addition, the president said he will ask Congress to legislate to ensure that American leadership on technology takes place in a context of “responsible innovation”.
The White House has not offered details on what the executive order will contain, but it will still have to take into account one aspect: namely how to control the ability of China and other competitor countries to advance faster than the US in artificial intelligence development programs.
One of the tracks that the White House follows is to restrict the export of advanced semiconductors and software to China. The G7 is aligned on the need to set rules for AI research to avoid what according to some philosophers and developers is an “inherent risk of extinction for humans” if AI takes over.
However, an executive order is a very clear step, which goes beyond the rules that the seven companies have voluntarily agreed to give themselves. After all, according to some experts, the list of standards is already part of the conduct that Silicon Valley companies have given themselves. Nick Clegg, president of global affairs at Meta, said he was “satisfied with these voluntary commitments entered into together” with the competitors. And the other CEOs have adapted along the same lines.
Safeguards include safety testing of AI products by independent experts and the sharing of product information with governments and entities working to manage technology risks. Another issue on which the seven companies have adapted is the guarantee for consumers to be able to recognize material produced by artificial intelligence; hence, the development of tools to address global and complex challenges such as disease and climate change. Furthermore, a qualifying point is that research and the potential of AI must take place without discrimination and without intrusion into privacy
Biden in his speech and presenting the agreement reiterated the need for innovation but this “cannot happen at the expense of the rights and security of Americans”.