Seeing is Believing? Deepfake Detection at the Time of Super Election Year
The year 2024 is shaping both technology and global politics. With the help of innovative AI products, it only takes a few clicks to create a text, image, or even a video that is shared on social media in a matter of moments, spreading the word for all to see. In addition to technological innovations, 2024 will be a super-election year, when 60 countries worldwide, including the USA and India, will hold national elections. More than 2 billion people worldwide are expected to vote, and political campaigns invest heavily to convince them.
What Impact Can We Expect From Technology to Influence Elections?
Spending on political advertising in the US is expected to reach a record $12.32 billion in 2024. The huge budget is also expected to have a big impact on society. That's why marketers need to understand the potential impact of their brand's positioning.
Social media first played an important role in 2007 during Barack Obama's election campaign, when his team pioneered the use of this technology to engage voters, raise funds and build a grassroots movement.
Meanwhile, the rise of social media, especially in political campaigns, has also created new threats, particularly the flow of misinformation and disinformation. By its very nature, social media is an information superhighway.
Regulating or recognizing misinformation becomes a challenge as such content strongly influences society in its decision-making.
The overlap of definitions and complexity makes things complicated. Misinformation refers to false or inaccurate information, regardless of intent.
On the other hand, disinformation is deliberately misleading information aimed at manipulating opinion. In particular, the spread of disinformation poses a significant threat to both political campaigns and brands.
Artificial intelligence and its innovations have two faces. It can be used both for innovative ideas and for the spread of misinformation and disinformation.
According to Forrester Research, 82% of marketers are aware of this threat and are seriously concerned about the impact of the presidential election cycle on their advertising efforts.
Be careful, the use of a curtain that changes a person's appearance or voice, as Tom Cruise did in the Mission Impossible films, is no longer just fiction.
Deepfakes refer to realistic audio or video recordings where people are seen saying or doing things they never did, meaning the face is swapped and used as a different person in the video.
Another application is attribute editing, where the person's features are changed, such as their style or hair colour. Face reenactment is also a typical deepfake, in which the facial expression is transferred to another person in a video.
The content can be formulated in such a way that it contains misleading information and becomes fake news. In this way, the entire text is falsified and the information is quickly passed on to other media outlets.
Fake reviews are already having a major impact on purchasing decisions on social media and in webshops, influencing the younger generation in particular in their purchasing behaviour.
Due to the speed at which content is generated and updated on social media, this is an extremely critical area for brands as posts, reels and stories spread quickly and generate false or misleading information. Bot-generated comments are already flooding social media and causing confusion or annoyance.
These few examples illustrate the many ways in which society, and brands in particular, can be harmed if they do not take proactive steps to protect themselves this year.
How Can Marketers Defend Their Reputation?
Marketers need to identify and effectively communicate their core value propositions and find their voice amongst the masses. The good old ‘save money, save energy’ messages are outdated. Serious discussions and issues arise during political campaigns, and a generic message can do even more damage and demonstrate a brand's lack of seriousness.
An articulated value proposition and consistent messaging across all channels reinforce brand identity and help consumers recognize misinformation. A strong and authentic brand that is known to the audience will be difficult for counterfeits to imitate, and even the most engaged audience will proactively report this information when they discover it.
As for brand content and assets, safe storage is required. Storing high-quality, original branded content in encrypted locations reduces the risk of unauthorized access and tampering. In addition, adding unique watermarks or digital signatures to branded content makes it easier to identify and track its authenticity.
When a deepfake occurs, a quick response must follow with accurate information and debunking efforts from the brand. For this reason, the entire team should be trained, not just in marketing but other departments as well, to understand what deepfakes are and the potential risks and damage they pose to the organization’s reputation.
Anyone who discovers a deepfake should report it to the brand team, who should take the necessary steps to get the situation under control.
Social media has a huge impact on the rapid spread of information, so this area needs to be monitored by highly trained professionals. Real-time tracking and brand monitoring should be carried out, with all brand mentions carefully scrutinized to identify deepfake content.
Adverts with a political reference are regulated on most social media channels, except TikTok. There is a labelling requirement for such ads if they use generative AI or synthetic content in their videos, images and audio.
Despite this labelling requirement, there are currently no government laws in place to regulate misinformation on social media or the ongoing impact of generative AI on advertising. So far, the European Union has begun formulating the EU AI law, which will come into force on 2 August 2026.
As a light on the horizon, Integral Ad Science unveiled a counterfeit detection technology in the summer that uses AI-powered multimedia analytics to accurately identify and mitigate harmful manipulations.
By combining this technology with existing misinformation categorization capabilities, marketers can protect their brands and ensure the authenticity of content in the face of this evolving threat. The beta version is available to select customers in the US and will hopefully be rolled out globally.
AI is an interesting technology that can be used to generate misinformation, but also to recognize fake content. It all depends on the input and the expected outcome.
Until the technology is fully mature enough to protect us from the threat of misleading information, humans should also be able to think more critically, ushering in the era of ‘seeing is not necessarily believing’.
Author:
Let’s make things #zSofisticated!
With a flair for creativity and a deep-rooted passion for digital marketing, Zsofia Raffa brings a unique perspective to the industry. Her extensive experience in crafting compelling campaigns for renowned brands such as Kodak Alaris, NVIDIA and Lufthansa has honed her ability to seamlessly bridge the gap between B2C and B2B realms.
A seasoned writer and journalist with a penchant for storytelling, Zsofia has contributed to numerous newspapers and publications. Her digital footprint extends beyond articles, encompassing engaging blogs, informative podcasts, and insightful newsletters.