Recently, under the promotion of the White House, Microsoft, Google, Meta, OpenAI, Amazon, Anthropic, an artificial intelligence startup founded by former OpenAI members, and Inflection, an artificial intelligence startup invested by NVIDIA and other companies, have joined the AI innovation.Both are committed to ensuring the security of artificial intelligence technologyCurrently, Anthropic, Google, Microsoft and OpenAI have jointly establishedIt is called the Frontier Model Forum.organizations to build safer and more responsible artificial intelligence computing models.
In previous discussions with US President Biden, seven technology companies investing in the development of artificial intelligence technology all pledged to abide by the AI safety principles, that is, to ensure that artificial intelligence is developed in a responsible manner to avoid negative impacts on public safety, rights and democratic values.
The newly established "Frontier Model Forum" is expected to establish an advisory committee, organizational charter and operating funds, and will also formulate core principles, which are expected to be completed within the next year. It will promote the research and development of AI technology security and work closely with policymakers, academic institutions, and private organizations to build AI technologies that benefit society.
The Frontier Model Forum also explains the qualifications for membership, including a commitment to the safety of the frontier technology models it develops and ensuring that artificial intelligence can bring greater social benefits.
In the consensus reached earlier between technology companies and the White House, they will ensure that their artificial intelligence can be tested by independent experts to see if there are any adverse effects. At the same time, they must continue to improve network security technology and encourage third-party companies, teams or individuals to report security vulnerabilities. In addition, they must ensure that artificial intelligence is not biased or used improperly, and conduct preliminary research on the possible social risks of artificial intelligence.
Other parts include the need to establish mutual trust mechanisms and share security information with other industries and government agencies, mark videos and audio content generated by artificial intelligence to facilitate easy identification by others, and agree to use cutting-edge artificial intelligence technology to solve social problems.
At the same time, the White House also stated that future artificial intelligence must comply with the three principles of safety, security and trust, and must be built in a responsible manner.
