Four months after issuing an executive order requiring artificial intelligence to be built and used safely, the Biden administration today announced the launch of the American AI Safety Institute Consortium (AISIC). The new consortium boasts more than 200 representatives, including top AI competitors Amazon, Google, Apple, Anthropic, Microsoft, OpenAI, and NVIDIA.
The consortium brings together AI developers, academics, government and industry researchers, civil society groups, and users to “develop and deploy secure and trustworthy artificial intelligence.”
“President Biden has directed us to use every tool at our disposal to achieve two key goals: setting safety standards and protecting the innovation ecosystem,” Commerce Secretary Gina Raimondo said in a statement. “This is what the National AI Safety Lab Consortium was established to help us.”
She explained that this consortium grew out of an executive order signed by U.S. President Biden last October. The mandate included developing guidelines for evaluating AI models, risk management, safety and security, and applying watermarks to AI-generated content.
“We will make sure America leads,” Raimando said. By working together with a group of leaders from industry, civil society, and academia, we can confront these challenges to develop the measurements and standards we need to sustain. “Secure America’s competitiveness and develop AI responsibly.”
The consortium includes representatives from the medical, academic, labor union, and banking sectors, as well as state and local governments, including JP Morgan, Citigroup, Carnegie Mellon University, Bank of America, Ohio State University, and Georgia Institute of Technology. representative.
Overseas partners are also expected to cooperate.
“This consortium represents the largest test and evaluation team established to date and will focus on establishing a new foundation of measurement science for AI safety,” the Commerce Department said. “The consortium… will work with organizations from like-minded countries that play a critical role in developing interoperable and effective tools for global safety.”
Since the list of participating companies is so extensive, it may be more useful to see which companies have not participated. Among the top 10 underrepresented technology companies are Tesla, Oracle, and Broadcom. TSMC is also not publicly traded, but it is also not a US-based company.
The rapid spread of generative AI tools into the mainstream has led to numerous instances of misuse and a surge in AI-generated deepfakes online. World leaders, including President Biden and former President Donald Trump, have been targeted by these fake images. On Thursday, the Federal Communications Commission announced that AI-generated robocalls using deepfake voices are illegal in the United States.
“These types of calls have increased over the past several years,” the FCC said, “because this technology now has the potential to confuse consumers with misinformation by imitating the voices of celebrities, political candidates, and close family members.” It was revealed.
Since the launch of GPT-4 early last year, world leaders have been considering how to govern AI development. Last May, the Biden administration met with several AI and technology companies, many of which are now part of the consortium. OpenAI, Google, Microsoft, Nvidia, Anthropic, Hugging Face, IBM, Stability AI, Amazon, Meta, and Inflection have all signed a pledge to develop AI responsibly.
“None of us can get AI right on our own,” Kent Walker, Google’s president of global operations, previously said. “We are excited to join other leading AI companies in supporting this commitment, and we are committed to continuing our collaboration by sharing information and best practices.”
editor: Ryan Ozawa.
