Please wait, loading...


The Year of Voting and Deepfakes | PR Ireland

Meanwhile back in the U.S., some voters in the State of New Hampshire received a phone call from President Biden telling them not to vote today. It was a very impressive fake.

The so called robocall is thought to have been generated by Artificial Intelligence. The U.S. Attorney General’s Office says it appears to be an unlawful attempt at voter suppression.

Audio deepfakes however are not just being made in the U.S. alone. Last December, there were more than 100 deepfake videos of the U.K. Prime Minister Rishi Sunak that someone had paid Facebook to promote.

Here in Ireland, the same mischief is beginning to take off with interested parties buying advertising on social media. In a world where these clips are everywhere, a question that we all ask is how do we limit the damage that these fakes can cause, particularly in this busy election year.

In researching this blog we have found a widespread mechanised impersonation campaign of the British Prime Minister. The motivating factors financially driven. Most of the images we discovered was to promote a financial scam to get people to hand over their money in the form of cryptocurrency.

Money they would never see again.

We found the Facebook campaign very interesting as they were all packaged up as fake news reports. It was a deepfake of the Prime Minister and deepfakes of BBC News presenters.

Here in Ireland we are seeing copycat campaigns of many famous faces such as Virgin TV News Colette Fitzpatrick, legendary RTÉ broadcaster Ann Doyle and An Taoiseach Leo Varadkar.

While it was financially motivated, political opinions and peoples perspectives of public figures may get caught in the cross fire.

This is really worrying at the moment. There are many elections all over the world this year. From the elections in the U.S. to the European local elections, with the overall election expected to be called in the months ahead. Irish Independent journalist Adrian Weckler showed online this week how easy it is to generate deep fake content.

These fake tools are now in everyone’s hands. These deepfake tools can create fake images and fake audio.

There is enormous scope for bad actors or bad state backed bad actors to affect elections in an enormous way. It is worth noting that there is potential for good things to happen as well. The biggest problem currently is that there are no real rules about how they should be used in elections.

There are no guidelines for campaigners, activists, political parties let alone bad actors operating on the dark web.

It would be good to see political parties and campaigners using generative AI technology to establish a public pledge to the voter and speak more publicly about responsible use of AI. It is worth noting that Facebook eventually took these deep fakes down and they continue to do so.

Looking to the years ahead we would expect to see accidental spreading of misinformation by some campaigners using generative AI technology and not fact checking it. The misinformation could occur by no one bothering to fact check or by simply not having a human in the loop. Misinformation can be peddled and spread and amplified by a simple accident.

AI has the potential to become a major disruptor in our domestic and global politics. As many of us know, misinformation has been around a long time and people have lied for political and personal financial gain.

The emergence of this technology can allow you to do it faster, bigger and better. As AI technology emerges and evolves, simple exercises will become of a higher quality. These audio recordings will sound more convincing as times goes on.

So with elections ahead of us, there is an urgent need of a legislative approach and seeking new rules to address the emergence of this new technology.

However, none of these pieces of legislation will be in place before the year end.

One safeguard that is being suggested to all AI companies and social media platforms are looking at ways to watermark material. Once again they will not be in place ahead of the election, instead we have to look at what is in our control ahead of the election. In our opinion all political parties should come together and agree a cross party approach to how they are going to use AI technology. There is an urgent lead to show leadership and start to set the norms about how to use it.

There is an urgent responsibility from everyone, from politicians and media, to citizens ourselves to be more skeptical of the media that we see.

There is a sensible small amount of paranoia here that we as citizens could have when we are consuming the media but there is also a huge role in the people creating media to use AI responsibly and safely and not let it run out of control and spread lies by accident.


Leave a Reply

Your email address will not be published. Required fields are marked *

Paul Allen & Associates Public RelationsContact Information
18 Fitzwilliam Square South, Dublin 2, D02 CY65, Ireland
GET IN TOUCHFollow us on
Paul Allen & Associates Public RelationsContact Information
18 Fitzwilliam Square South Dublin 2 Ireland
OUR LOCATIONSWhere to find us?
img footer map
GET IN TOUCHFollow us on
2020 © Paul Allen PR & Associates. All rights reserved.