GW will lead the new $20 million NSF AI agency

WASHINGTON (May 16, 2023) – The George Washington University is co-leading a multi-agency effort supported by the National Science Foundation (NSF) that will develop new artificial intelligence (AI) technologies designed to increase trust and reduce risk, while empowering and educating communities.

The NSF Institute for Trustworthy AI in Law & Society (TRAILS) announced on May 4, 2023, bringing together AI and machine learning specialists with systems engineers, social scientists, legal scholars, educators, and public policy experts. The multidisciplinary team will work with affected communities, private industry, and the federal government to determine how to evaluate trust in AI, how to develop technical solutions and processes for trustworthy AI, and which policy models best create and maintain trust.

“TRAILS Institute is a prime example of the world-class work that GW faculty is doing at the intersection of technology and the social sciences, and it reflects the university’s commitment to innovation that advances knowledge, boosts the economy, and improves educational equity and quality. life for all members of our global community,” said GW President Mark Wrighton.

GW Chancellor and Executive Vice President of Academic Affairs, Christopher Bracey, echoed the President’s excitement about this new collaboration. “With engineers and computer scientists working side by side with legal scholars and social scientists in the policy capitals of the world, TRAILS is poised to increase the opportunities and address the risks of artificial intelligence and how it is used in our society. ”

Funded by a $20 million award from NSF, the new agency hopes to transform AI practices by driving new innovations that promote ethics, human rights, and input and feedback from communities whose voices were previously marginalized.

John Lach, Dean of the GW School of Engineering and Applied Science, said, “TRAILS embodies the ‘Engineering And…’ ethos of GW Engineering – working across disciplines and engaging with communities to harness the power of engineering and computing for the good of society.”

In addition to GW, TRAILS will include faculty members from the University of Maryland, including Hal Daumé III, principal investigator for TRAILS, Morgan State University, and Cornell University, with more support coming from the National Institute of Standards and Technology (NIST), and other organizations. private sector such as Arthur AI, Checkstep, FinRegLab, and Techstars.

The new institute recognizes that AI is currently at a crossroads. AI-infused systems have great potential to increase human capacity, increase productivity, catalyze innovation, and mitigate complex problems, but current systems are developed and deployed in processes that are closed to the public. As a result, those most affected by technology have little to say about how it was developed.

“I think it’s important to understand that AI can be a significant source of benefit and innovation for society, but it can also cause a lot of harm,” said David Broniatowski, a professor of engineering management and systems engineering at GW and principal investigator of TRAILS at GW. “Many of those losses are felt by people who have been historically underrepresented because their concerns were not reflected in the design process.”

For example, an AI system can be trained on a data set that reflects the values—and biases—of the system designer or data labeler. People using the system may not be aware of the bias, believing instead that the output of the system is “objective”. If system outputs are not easy to interpret, the priorities encoded into the system may reflect values ​​that are inconsistent with those using, or affected by, the system, Broniatowski said.

Given these conditions—and the fact that AI is increasingly being used in systems that are critical to society, such as mediating online communications, determining healthcare choices, and offering guidance within the criminal justice system—there has been an urgency to ensure that societal trust in AI systems matches the level of same system trust.

TRAILS has identified four main research drives to promote the development of AI systems that can gain public trust through wider participation in the AI ​​ecosystem.

The first, known as participatory AI, supports the involvement of human stakeholders in the development, implementation and use of these systems. It aims to create technology in a way that aligns with the values ​​and interests of different groups of people, rather than being controlled by a few experts or driven purely by profit.

The second research thrust will focus on developing sophisticated machine learning algorithms that reflect the values ​​and interests of relevant stakeholders.

Broniatowski will lead research by the three institutes to evaluate how people perceive the developed AI systems, and the degree to which their levels of reliability, fairness, transparency and accountability will lead to appropriate levels of trust.

Susan Ariel Aaronson, a research professor of international affairs at GW, will use her expertise in data-driven change and international data governance to lead the four institutes’ push on participatory governance and trust.

“There is no trust without participation and no accountability without participation – hence we believe in a participatory approach to AI at all levels from design to deployment,” said Aaronson.

Morgan State University, Maryland’s leading urban research university, will lead community-based projects related to the interaction between AI and education, while Cornell University will advance efforts focused on how people interpret their use of AI.

Federal officials at NIST will collaborate with TRAILS on the development of meaningful measures, benchmarks, test beds, and certification methods—especially as they apply critical topics critical to trust and confidence such as security, fairness, privacy, transparency, explanation, accountability, accuracy. and reliability.

“The ability to measure the trustworthiness of AI systems and their impact on individuals, communities and society is limited. TRAILS can help advance our understanding of the fundamentals of trustworthy AI, the ethical and social considerations of AI, and how to build systems that are trusted by the people who use and are affected by them,” said Under Secretary of Commerce for Standards and Technology and Director of NIST Laurie E. Locascio.

Today’s announcement is the latest in a series of federal grants that make up the National Artificial Intelligence Research Institute group. Recent investments in seven new AI agencies, totaling $140 million, follow on from the previous two award rounds.

NSF, in partnership with government agencies and private sector leaders, has now invested nearly half a billion dollars in the AI ​​institute ecosystem—investments that extend collaborative AI research networks to nearly every US state.

“This institution drives breakthrough discoveries to achieve our country’s ambition to be at the forefront of the global AI revolution. This work would not have been possible without our long standing alliances with academic partners, government agencies, industry leaders and the AI ​​community at large,” said NSF Director Sethuraman Panchanathan.

###

This research was supported by the National Science Foundation (Award IIS-2229885). This story does not necessarily reflect the views of this organization.

Media Contacts: (GW) Caitlin Douglass, (email protected); (NSF) 703.292.7090, (email protected)

About the National Science Foundation

The US National Science Foundation is an independent federal agency supporting science and engineering in all 50 US states and territories. NSF was founded in 1950 by Congress to promote scientific progress; promote national health, prosperity and welfare; and secure national defense. The organization fulfills its mission mainly by giving grants. NSF investments account for approximately 25% of federal support to American colleges and universities for basic research: research driven by curiosity and discovery. NSF also supports solution-oriented research with the potential to make progress for the American people.

About George Washington University

Chartered on February 9, 1821, and located in the heart of Washington, DC, George Washington University provides unparalleled access to leading international institutions, multinational corporations, global media outlets, and governments from 177 countries through their embassies. This is a singular advantage—no other university has this much potential for international engagement in their stride.