Conference: Workshop on Designing Accountable Software Systems

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

Society is increasingly dependent on software applications, systems, and platforms, as functionality in all aspects of business, government, and everyday life is becoming implemented in software-intensive systems. Whereas people and enterprises have been held accountable to laws, regulations, and societal goals and norms, when implemented through software, now software systems must also work in a way consistent with accountability goals. Software systems need to be designed with legal and regulatory compliance and social norms in mind. The U.S. National Science Foundation (NSF) Designing Accountable Software Systems (DASS) program is supporting research investigating the issues in achieving accountable software systems. Researchers on DASS proposals are a mix of Software Engineering researchers paired with PIs from the social and legal science. The purpose of the DASS workshop is to assess the state of the DASS program and the state of related research at large. In addition to reflecting on the successes and challenges of DASS-related research from the past, the workshop will focus on the emergence and expansion of DASS-like issues in the context of established research fields, such as safety in cyber-physical systems, fairness in artificial intelligence (AI) systems, and privacy and security in cyberspace. The workshop will help grow the DASS community of researchers and provide valuable insights into existing research challenges and future research directions in the field. By improving accountability in software design, the success of building this community can improve how members of society from across socio-cultural and economic backgrounds engage with services across banking, education, healthcare, and transportation, among others. Designing accountable software systems relies on theory and scientific principles from across the social and behavioral sciences, law, and computer science, including how individuals and groups make decisions, how bias affects decision-making, and how incentives motivate individuals. Legal theory that differentiates compliance from liability and legal processes that produce laws and adjudicate whether individuals and groups are compliant with law all affect how we define accountability. Finally, the technical means to support decision-makers and implement processes must be demonstrably correct and reliable. The software development processes used in practice must take these advances into account. A key challenge is how to nurture collaboration across these varied disciplines. The workshop will create an environment where ideas are presented, developed, discussed and synthesized to orient this cross-disciplinary community around a shared set of scientific objectives, each of which is originated and developed by one or more disciplines and supported by others. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.