Image was generated by the AI tool Midjourney MIDJOURNEY/REDDIT
Image was generated by the AI tool Midjourney MIDJOURNEY/REDDIT

BACKGROUND BRIEFING: Deepfakes are here to stay - so where to next?

Embargoed until: Publicly released:
Not peer-reviewed: This work has not been scrutinised by independent experts, or the story does not contain research data to review (for example an opinion piece). If you are reporting on research that has yet to go through peer-review (eg. conference abstracts and preprints) be aware that the findings can change during the peer review process.

****RECORDING AVAILABLE*****From fake audio impersonating politicians to fake pornographic videos of celebrities including Taylor Swift, deepfakes have been in the news a lot lately and advances in artificial intelligence (AI) mean it's about to get a whole lot worse, say experts. So, what are the biggest challenges and what can people do to avoid being duped by deepfakes and images and video generated by AI? Join us for this online briefing to discuss the latest trends in AI deepfake technology and potential solutions.

Organisation/s: Australian Science Media Centre, The University of Adelaide, University of Technology Sydney (UTS), The University of Sydney

Funder: This briefing includes speakers from organisations that are financial supporters of the AusSMC. A full list of supporters is available at smc.org.au

Media release

From: Australian Science Media Centre

BACKGROUND BRIEFING: Deepfakes are here to stay - so where to next?

Wednesday 21 February 2024 at 10am AEDT ONLINE

From fake audio impersonating politicians to fake pornographic videos of celebrities including Taylor Swift, deepfakes have been in the news a lot lately and advances in artificial intelligence (AI) mean it's about to get a whole lot worse, say experts.

So, what are the biggest challenges and what can people do to avoid being duped by deepfakes and images and video generated by AI, especially as the technology is set to become even more sophisticated with the development of Sora, an AI that can generate videos from text prompts?

Is there a technological solution and how far away is it? Can Australia rely on overseas technology or do we need to develop our own capability?

While it may seem that the horse has bolted on deepfakes, what more can governments and industry do to manage the growing avalanche of fake AI-generated content?

Join us for this online briefing to discuss the latest trends in AI deepfake technology and potential solutions.

The background briefing will discuss topics such as:

  • The current state of generative AI image technologies
  • The challenges of distinguishing between authentic content and convincing AI content
  • The potential for AI to both harm and enhance journalistic integrity
  • The necessary steps for media professionals, governments, regulators, and the broader public to navigate this rapidly evolving landscape in 2024
  • Deepfakes are one face of a very complex issue. Navigating the underlying ethical concerns behind generative AI can help us better prepare for responsible adoption of the technology.

Speakers:

  • Will Berryman, Executive Director of the Royal Institution of Australia (RiAus)
  • Professor Simon Lucey, Director of the Australian Institute for Machine Learning (AIML) at The University of Adelaide
  • Professor Monica Attard, Co-Director of the Centre For Media Transition at The University of Technology Sydney
  • Rebecca Johnson, AI and ethics researcher, The University of Sydney (Rebecca is the final year of a PhD on ethics in AI - read her full bio at ethicsgenai.com)

Date: Wednesday 21 February 2024
Start Time: 10:00am AEDT
Duration: Approx 45 min 
Venue: Online - Zoom

Attachments:

Note: Not all attachments are visible to the general public

News for:

Australia
NSW
SA

Media contact details for this story are only visible to registered journalists.