Partnership on AI 2019 Annual Report: Building a Connected Community for Responsible AI

The following is an abbreviated excerpt from PAI’s 2019 Annual Report – Letter from the Founding Executive Director

The Partnership on AI was founded two years ago with a simple premise: AI is already having a transformative impact on our society, and the essential work of ensuring that AI technologies are fair, safe, just, and inclusive is a critical responsibility that must be collectively prioritized.

PAI believes that the AI developer community has a special obligation to push for, understand, and ultimately implement changes around the design and practice of ethical technology. We also believe that this work must be informed by diverse voices, multidisciplinary expertise, and a deep understanding of the context in which technology is being used.

Our first Annual Report (2019) documents how the power of this premise has come to life in the Partnership’s work. Over the last year, PAI has been proud to emerge as a hub of experimentation in multistakeholder design, and a place where competing entities and diverse equities can converge to foster shared dialogue and identify a common cause. We have tackled urgent topics that require shared understanding, more effective coordination, or deeper investment and attention from the global community. These include the impacts of AI on media manipulation, the use of algorithmic assessments in high-stakes domains, how to improve transparency and explainability in machine learning systems, and freedom of movement and scientific collaboration for the global AI community—among others.

In 2019, we also grew our community to over one hundred Partner organizations, representing a diverse array of interests, spanning 13 countries and four continents. We hosted our second-annual All Partner Meeting in London, where over two hundred community members gathered to advance existing projects, talk about the most urgent needs of the field, and to launch new work. Notably, we added 21 new team members to our growing staff. We gathered PAI community members in Montgomery, San Francisco, London, New York, Atlanta, Boston, Berlin, Beijing, Luxembourg, Montreal, Chicago, Seattle, Vancouver, Philadelphia, and more. We launched the PAI Fellowship Program and ended the year with an impressive research publication record in several major research conferences.

As we look toward this next year of our evolution, the Partnership aims to drive a race-to-the-top in moral leadership among the leading entities and individuals developing AI. This work could not come at a more crucial time: the global community requires an evidence base for the design, practice, and governance of responsible AI. We will continue to support practitioners of AI in not shying away from the hard questions, provide tools for advocates protecting essential rights, and to work to ensure that institutions and individuals empowered to make decisions about the future of technology development do not cede their critical responsibilities to society. We will strive to develop the AI field such that it is deeply informed by other disciplines, and by front-line communities impacted by technology every day. And we will endeavor to share what lessons about inclusive dialogue we learn along the way. Download our 2019 Annual Report for more details.

I look forward to continuing this important work hand-in-hand with our community as we head into 2020 and beyond.