Today, Microsoft is declaring its endorsement of the new voluntary pledges established by the Biden-Harris administration to guarantee the safety, security, and reliability of advanced AI systems. By backing all the voluntary pledges introduced by President Biden and also making additional independent commitments that align with these essential objectives, Microsoft is enhancing its practices in safe, responsible, and trustworthy AI, collaborating with fellow industry frontrunners.
The White House's swift actions lay the groundwork for safeguarding the potential of AI by staying ahead of its potential risks. We appreciate the President's initiative in uniting the tech sector to establish practical measures that enhance the safety, security, and public benefits of AI.
Guided by the enduring principles of safety, security, and trust, the voluntary commitments aim to address the challenges posed by advanced AI models and encourage the implementation of specific practices – such as red-team testing and the publication of transparency reports – that will drive progress across the entire ecosystem. These commitments build on the solid groundwork laid by the U.S. Government (including the NIST AI Risk Management Framework and the Blueprint for an AI Bill of Rights) and complement the existing measures for high-risk applications in Europe and other regions. We anticipate widespread adoption of these commitments by the industry and their integration into the ongoing global dialogues regarding the formulation of an effective international code of conduct.
Microsoft is significantly strengthening its commitments to fortify the ecosystem and uphold the core values of safety, security, and trust. By championing a trial of the National AI Research Resource and advocating for the establishment of a national registry for high-risk AI systems, we firmly believe that these initiatives will foster transparency and accountability. Moreover, we are wholeheartedly committed to extensively implementing the NIST AI Risk Management Framework and adopting cybersecurity protocols customized to address specific AI risks. We are confident that these proactive measures will lead to the development of more dependable AI systems that deliver advantages not only to our clients but to society at large.
You can view the detailed commitments Microsoft has made here.Â
Effective collaboration within a team is crucial for shaping and executing commitments, such as those at Microsoft. I wish to extend my appreciation to Kevin Scott, the Chief Technology Officer at Microsoft, with whom I collaborate closely to oversee our responsible AI program. I also wish to recognize the invaluable contributions of Natasha Crampton, Sarah Bird, Eric Horvitz, Hanna Wallach, and Ece Kamar, who have played a pivotal role in guiding our responsible AI ecosystem.
Similar to the voluntary commitments outlined by the White House, our AI initiatives should always prioritize people. I am grateful for the unwavering support of Microsoft's strong leadership team, which enables us to honor our commitments and advance the program we have diligently developed over the past seven years. By establishing ethical standards early in the trajectory of this technology, we not only ensure its safety, security, and reliability but also enhance our capacity to utilize AI for positive impacts in communities both within the U.S. and globally.