Letting the Machine Decide? Privacy Risks Associated With AI Technologies

No ratings

Presented at GlobalPrivacySummit 2019 by

Ruby Zefo

Alexis Goltra

Christine Lyon

From scouring applicants’ resumes to detecting insider threats, AI technology is already revolutionizing the workplace. In the consumer context, companies are harnessing AI to maximize marketing and other customer-targeting initiatives and detect fraud. However, AI’s seemingly limitless potential is not without privacy and reputational risks. This panel will address the risks—several of which have become even more critical in light of the GDPR’s enforceability and the recent passage of the California Consumer Privacy Act—and ways companies can mitigate them in developing and implementing AI technologies. For instance, how will companies address notice and transparency requirements, data subject access and deletion rights, and restrictions on profiling and automated decision-making? How will they prevent algorithmic bias, wherein the implicit biases of the humans who “train” the algorithms are reflected in their outputs—an issue that has captured news headlines and threatens legal, financial and reputational harm?What you’ll take away:An understanding of the various privacy principles imperiled by AI technologies“The algorithm did it” defense is inadequate to shield companies from liability stemming from unintended or undesirable outcomesRegulatory developments, including the passage of the California Consumer Privacy Act, signal that these issues are not going away; to the contrary, they are more important than ever