Scottish Parliament TV could become a target for deepfakes says new research

Scotland was one of the world’s first parliaments to introduce livestreaming from its Debating Chamber and committee rooms with hours of footage being generated many weeks which, because of sophisticated AI technologies, could now be misused to create convincing deepfakes according to new research published today (Tuesday 10 December 2024).

While deceptive videos and audio clips could pose a serious threat to Scottish democracy, the good news is researchers based at the Scottish Centre for Crime and Justice Research and the University of Edinburgh have found the parliament is resilient against these kinds of attacks.

The team was commissioned by the Scottish Parliament to evaluate the potential threat of Artificial Intelligence (AI) technologies to parliamentary businesses and to consider how deepfake videos may affect the integrity of Parliament TV, which one participant dubbed ‘chamberfakes’.

Despite finding many examples of good practice, the team are calling on the Scottish Parliament to go further and put a formal process in place to respond to deepfake threats and allocate more staff to manage the risks of attacks.

Deepfakes are created when a video, image, or audio file has been changed using AI to distort someone’s words or actions. Due to recent advancements in technology, it has now become easier to create convincing deepfakes.

The research team identified three main risks including the video livestream being hacked, the dissemination of deepfakes on social media, and the many hours of parliamentary footage being used as feedstock to create harmful and abusive deepfakes of MSP’s.

The findings have been published by the Scottish Centre for Crime and Justice Research in a report called ‘Chamberfakes: assessing the threats posed by generative AI technologies to parliamentary democracy in Scotland.’

Co-author of the report, Dr Ben Collier who is based at the Scottish Centre for Crime and Justice Research (SCCJR) and the University of Edinburgh, said: “Scotland was one of the world’s first parliaments to experiment with live streaming. While the original aim was to produce a neutral record for the public and broadcasters on what goes on inside Holyrood, it paradoxically has introduced new security risks that could threaten public trust in democracy.

“While the risk of a widespread attack is technically low, we have seen one-off examples of Scottish politicians falling victim to these kinds of attacks such as Maggie Chapman MSP, who was the subject of a deepfake audio clip posted on X that parodied Chamber business.

“The ramifications of ‘chamberfakes’ are potentially severe but we are optimistic about the Scottish Parliament’s ability to respond to these sorts of attacks in large part due to the skilled broadcasting team who have technical processes and human checks already in place.”

The authors of the report believe there is always room for improvement which is why they include three key recommendations to bolster the existing checks and procedures.

Co-author Dr Morgan Currie commented: “At the moment the Scottish Parliament has no formal processes in place to respond to deepfake threats. We recommend a formal intervention plan be put in place and reporting procedure for deepfakes and misinformation introduced and managed by a specific member of staff.

“Consideration should also be given to developing simple ways of authenticating witnesses and other contributors who dial in remotely to give evidence to parliamentary committees.

“And our final recommendation is for a new team to be established within the broadcasting unit at the parliament to support those MSPs who encounter or are victims of misinformation who could track how that content is being used while also promoting the use of parliamentary video through communications campaigns and direct engagement with broadcasters and social media platforms.”

END

Notes to Editors:

1. The Scottish Centre for Crime and Justice Research (SCCJR) is a collaboration between the Universities of Glasgow, Edinburgh, Edinburgh Napier, Stirling and Strathclyde. SCCJR aims to produce research that informs policy and practice and advances our understanding of justice.
2. ‘Chamberfakes: assessing the threats posed by generative AI technologies to parliamentary democracy in Scotland’ was published on 10 Dec 2024 and is available on the SCCJR website. The report was written by Dr Ben Collier, Dr Morgan Currie and Dr Benedetta Catanzariti.
3. This report is based on a review of current academic and policy literature on political deepfakes. The team supplemented secondary sources with 17 interviews, including eight staff from the Scottish Parliament, along with people working in journalism, broadcasting, academia, cybersecurity and video streaming, and one Member of the Scottish Parliament. The interviews took place July through October 2024. The research team also made site visits to the Scottish Parliament to observe the video team on days when the Parliament was in session to trace the technical chain, practices, and procedures that come together to produce parliamentary business.
4. The research was commissioned by the Scottish Parliament.