The University of Texas at Austin and the MITRE Corporation, a nonprofit dedicated to solving problems for a safer world, have formed a partnership that includes accelerating innovative ethical artificial intelligence (AI) research currently underway by interdisciplinary teams of researchers who are part of UT Austin's Good Systems research grand challenge.

The partnership will involve Peter Stone and Justin Hart from UT Computer Science, as well as other experts from the Cockrell School of Engineering, the School of Information, and the Department of Defense Research Advancement.

AI technologies are being adopted rapidly across a wide range of industries due to the increased efficiency and convenience they can provide. However, the widespread adoption of some AI technologies can cause adverse societal effects, among them job losses, increased economic inequality, infringements on privacy, overreliance on biased algorithms and data sets, and the disruption of elections and of the social fabric.

"Good Systems is honored to partner with MITRE to work toward our shared goal of ensuring that AI will serve the public interest," said Kenneth R. Fleischmann, founding chair of Good Systems and a professor in the School of Information. "AI innovation should be leveraged to enhance our economic competitiveness and national security, and it is also critical that we harness the benefits of AI to steer toward a more equitable and just society."

In 2019, the university launched Good Systems, one of three Bridging Barriers research grand challenges that bring together researchers from dozens of disciplines to address major issues affecting our world.

Since then, technologists, social scientists and humanists participating in Good Systems have been working to define, evaluate and build ethical AI systems in six core areas:

• Enhancing middle-skill work with smart hand tools
• Designing responsible AI to curb disinformation
• Developing AI systems for smarter cities
• Embedding ethics in public camera AI technologies
• Designing AI to advance racial equity
• Living and working with robots

MITRE's investment of $1 million in support of Good Systems will allow investment in capacity building in these core areas by expanding its team expertise and supporting a greater number of faculty members, researchers and students to engage in this work.

This gift will serve as a model for an ongoing flagship partnership with UT Austin and will be coordinated by Defense Research Advancement in the Office of the Vice President for Research, Scholarship and Creative Endeavors.

The collaboration between Good Systems and MITRE was made possible through the efforts of the UT Austin Cockrell School of Engineering, College of Natural Sciences, School of Information and Department of Defense Research Advancement, including facilitating a MITRE site visit to UT Austin during February.

In addition, Mikel Rodriguez, the director of MITRE's Artificial Intelligence and Autonomy Innovation Center, was an invited speaker at the Good Systems Symposium in April. Fleischmann will be an invited speaker at MITRE's Grand Challenges Power Hour: Building Trust – AI Assurance in Industries & Mission Critical Systems in July.

"As consequential use of AI increases, it is vital that we address safety, security and equity concerns," said Douglas Robbins, vice president of engineering and prototyping at MITRE Labs. "We are thrilled to be working with UT Austin and Good Systems to advance the underlying science in these areas that will enable us to build AI systems that can reach their full potential."

Cross-posted from the College of Natural Sciences News