Our commitments to advance protected, safe, and reliable AI

Today, Microsoft is asserting its assist for brand new voluntary commitments crafted by the Biden-Harris administration to assist be sure that superior AI programs are protected, safe, and reliable. By endorsing the entire voluntary commitments introduced by President Biden and independently committing to a number of others that assist these essential targets, Microsoft is increasing its protected and accountable AI practices, working alongside different business leaders. 

By shifting shortly, the White House’s commitments create a basis to assist make sure the promise of AI stays forward of its dangers. We welcome the President’s management in bringing the tech business collectively to hammer out concrete steps that can assist make AI safer, safer, and extra helpful for the general public. 

Guided by the enduring rules of security, safety, and belief, the voluntary commitments handle the dangers introduced by superior AI fashions and promote the adoption of particular practices – similar to red-team testing and the publication of transparency reviews – that can propel the entire ecosystem ahead. The commitments construct upon robust pre-existing work by the U.S. Government (such because the NIST AI Risk Management Framework and the Blueprint for an AI Bill of Rights) and are a pure complement to the measures which have been developed for high-risk functions in Europe and elsewhere. We stay up for their broad adoption by business and inclusion within the ongoing world discussions about what an efficient worldwide code of conduct may appear like. 

Microsoft’s extra commitments give attention to how we are going to additional strengthen the ecosystem and operationalize the rules of security, safety, and belief. From supporting a pilot of the National AI Research Resource to advocating for the institution of a nationwide registry of high-risk AI programs, we consider that these measures will assist advance transparency and accountability. We have additionally dedicated to broad-scale implementation of the NIST AI Risk Management Framework, and adoption of cybersecurity practices which might be attuned to distinctive AI dangers. We know that it will result in extra reliable AI programs that profit not solely our clients, however the entire of society. 

You can view the detailed commitments Microsoft has made right here. 

It takes a village to craft commitments similar to these and put them into follow at Microsoft. I want to take this chance to thank Kevin Scott, Microsoft’s Chief Technology Officer, with whom I co-sponsor our accountable AI program, in addition to Natasha Crampton, Sarah Bird, Eric Horvitz, Hanna Wallach, and Ece Kamar, who’ve performed key management roles in our accountable AI ecosystem.  

As the White House’s voluntary commitments mirror, folks should stay on the middle of our AI efforts and I’m grateful to have robust management in place at Microsoft to assist us ship on our commitments and proceed to develop this system we now have been constructing for the final seven years. Establishing codes of conduct early within the improvement of this rising know-how won’t solely assist guarantee security, safety, and trustworthiness, it’ll additionally permit us to raised unlock AI’s constructive impression for communities throughout the U.S. and world wide.