Deepfakes are the hot topic in the media today. But what are they and why does IFLA need to be concerned with this emerging technology?

Deepfakes are manipulated audio-visual content created by machine-learning algorithms. For example, a person in an existing image can be replaced with another’s likeness and that likeness can be made to say and do things the person never did. The tools to create and disseminate this new form of manipulated media are easier, faster, cheaper, and more accessible than ever, and today convincing deepfakes can be produced in under 24 hours for less than $1000. The challenge now is not just to raise awareness, but also to propose and develop effective responses. Now that deepfakes are here, what do we do about this phenomenon?

This session will provide an overview of the technology and its implications for social trust, discuss examples of deepfakes in entertainment, politics, financial markets, and privacy, and analyze why this problem is so difficult to solve. Finally, it will address the policy, technology, and education solutions required to address malicious deepfakes.

View the recorded lecture here.


Suggested Audience: Librarians, publishers, general audience of interested parties

Organiser: Ann Okerson ([email protected])

Sponsored by: IFLA News Media Section