Regulating Machine Learning

Cary Coglianese’s recent paper, “Regulating Machine Learning: The Challenge of Heterogeneity,” published by Competition Policy International, argues that because machine-learning algorithms are so dissimilar and diverse—that is, they are heterogeneous—figuring out the best way to regulate their use will be challenging.

Law
Political Science

Social Norms for Algorithms

Responsible AI in the generative era

Generative AI raises new challenges in defining, measuring, and mitigating concerns about fairness, toxicity, and intellectual property, among other things. But work has started on the solutions, writes Michael Kearns for Amazon Science.

Computer and Information Science

Social Norms for Algorithms

U.S. Census Data Vulnerable to Attack Without Enhanced Privacy Measures

With computing power no stronger than a commercial laptop, Aaron Roth and Michael Kearns recently demonstrated that statistics released by the U.S. Census Bureau can be reverse engineered to reveal protected info about individual respondents.

Computer and Information Science

Social Norms for Algorithms

Wharton Researchers Build Algorithm to Help Fight Sex Trafficking

Hamsa Bastani and Pia Ramchandani are using machine learning to shine a light on sex trafficking: they spent countless hours running trials, examining the results, and redesigning the algorithms to get everything right. The ultimate goal, they said, was to “train” the algorithms to hit the right target through the right keywords.

Operations Information and Decisions

Social Norms for Algorithms

Service Cards and ML Governance

Michael Kearns joined the TWIML AI Podcast to discuss responsible AI, ML governance, the announcement of service cards, and their take on “model cards” at a holistic, system level as opposed to an individual model level.

Computer and Information Science

Social Norms for Algorithms

Keys to Getting the Most From Machine Learning

To maximize results from machine learning and artificial intelligence, you need to tailor your tech to the task at hand. Hamsa Bastani explores solutions for making the most of AI in both her undergraduate Introduction to Management Science course and lectures for Executive Education.

Operations Information and Decisions

Social Norms for Algorithms

Racial Equity and Data Privacy

In this episode of the Race and Regulation Podcast, Anita Allen reveals how race-neutral privacy laws in the U.S. have failed to address the unequal burdens faced online by Black Americans, whose personal data are used in racially discriminatory ways.

Law
Philosophy

Social Norms for Algorithms

American Philosophical Society Membership

Anita Allen has been elected to the prestigious American Philosophical Society. The Society was founded in 1743 by Benjamin Franklin for the purpose of “promoting useful knowledge” and elects fewer than 30 resident members annually.

Law
Philosophy

Social Norms for Algorithms

At the Duke Law Journal, Cary Coglianese offers a framework for determining when government should use artificial intelligence

In a pathbreaking article recently published in the Duke Law Journal, Cary Coglianese and Alicia Lai L’21 explore governmental reliance on digital algorithms, concluding that “public officials should proceed with care on a case-by-case basis” when deciding whether to employ digital algorithms in place of “human algorithms.”

Law
Political Science

Social Norms for Algorithms

Using Machine Learning To Combat Human Trafficking

Hamsa Bastani discusses key findings from her research on how deep web data and machine learning can help law enforcement agencies investigate and intervene in human trafficking in Wharton Social Impact’s “Research Spotlight” series.

Operations Information and Decisions

Social Norms for Algorithms

News

Regulating Machine Learning

Responsible AI in the generative era

U.S. Census Data Vulnerable to Attack Without Enhanced Privacy Measures

Wharton Researchers Build Algorithm to Help Fight Sex Trafficking

Service Cards and ML Governance

Keys to Getting the Most From Machine Learning

Racial Equity and Data Privacy

American Philosophical Society Membership

At the Duke Law Journal, Cary Coglianese offers a framework for determining when government should use artificial intelligence

Using Machine Learning To Combat Human Trafficking