The use of large language models is creating data privacy issues because they tend to disclose personal information they scrape from the internet without the consent of the data subject, Axios reports. Artificial intelligence data leaks come in many forms, such as an accidental disclosure or through malicious actions like building a model that circumvents privacy controls. Editor's note: Explore the IAPP AI Governance Center and subscribe to the AI Governance Dashboard.
14 March 2024
Large language models present privacy risks
RELATED STORIES
A view from DC: Our data, ourselves
Notes from the IAPP Canada: Privacy law teaching, training and turkey
Notes from the Asia-Pacific region: Quiet, contemplation follows introduction of initial Australia Privacy Act reforms
A view from Brussels: Trans-Atlantic transfers deal gets thumbs-up as adequacy remains a strategic priority
US courts, regulators weigh in on online tracking in health care