AAAIP Gives Testimony in Strong Support of Massachusetts Bill Targeting AI Election Interference
If passed, bill S.2730 would ban the distribution of deceptive or fraudulent synthetic media messages which depict a candidate or political party set to appear on the state or local ballot within 90 days of an election. This week, Dr. Hamid Ekbia provided testimony supporting the bill on behalf of AAAIP.
Read the testimony below:
Dear Members of the Committee:
Good morning,
I am a professor and director of an AI institute at Syracuse University in New York State. Today, I represent the academic alliance for AI Policy that currently involves around 40 major universities in the country. Our members are academics with expertise in computer science, engineering, and AI, as well as social science, law, and policy. Our members are in strong support of S.2730, a legislative proposal that seeks to regulate deepfakes in election communications.
Recent developments in Generative AI and reduced barriers to entry has given rise to an increasingly volume of deepfakes and other forms of synthetic content with significant social and political implications — from the saturation of false information about a particular topic or event to loss of common understanding and trust in social institutions. Moreover, affective polarization and negative partisanship, along with associated psychological mechanisms such as confirmation bias and motivated reasoning, push individuals toward misinformation-promoting environments.
These are real and damaging threats to democracy. The Academic Alliance for AI Policy strongly urges the Joint Committee on Election Laws to move S.2730 forward to reduce the potential harms of these social and psychological mechanisms by regulating deepfakes in election communication. This is hopefully a first step in protecting our citizens and our institutions from the attacks and agendas of maleficent players.
Thank you again for the opportunity to testify in support of S.2730. I would be happy to answer any questions.