Praveen Gujar, director of product at LinkedIn, has 15+ years' experience launching enterprise products in digital advertising and AI/ML.
getty
Breakthroughs in technology often invite new types of misuse. Generative AI (GenAI), for example, has transformed content creation, but it has also made it easier and more cost-effective to produce deepfakes.
This year, the rise of deepfakes is particularly troubling. With more than four billion people eligible to vote in more than 40 countries in 2024, deepfakes and deepfake ads can be prime vectors of civic abuse.
To combat this, we need a multifaceted strategy across technology, regulations and education.
Good, Bad And Ugly Of Deepfakes
Deepfakes blend AI with the art of fabrication by utilizing generative adversarial networks (GANs) to craft videos and audio that impersonate people saying or doing things they've never actually said or done.
Deepfakes unlock remarkable possibilities, from crafting lifelike digital avatars in films and games to innovative educational use cases—like the 2019 global malaria campaign where David Beckham communicated in multiple languages.
On the flip side, deepfakes can be used to spread misinformation, tarnish reputations or perpetrate fraud, such as manipulated ads featuring Joe Rogan endorsing a supplement he didn't endorse. Public figures such as politicians and celebrities are particularly vulnerable victims.
The surge in computing power coupled with affordable cloud computing, the widespread availability of data and the prowess of large language models (LLMs) have simplified and reduced the cost of producing deepfake content. However, regulations and content moderation have lagged behind—despite new acts like the European AI Act and coalitions like C2PA—fueling barrages of forged content.
Deepfake (Ads) And Civic Harm
With social media, deepfakes can easily achieve viral or widespread distribution. Deepfakes often elicit strong emotional reactions, whether through shock, humor or outrage. Social media algorithms can then amplify such messages, enabling the content to spread exponentially unless flagged by users or systems. This may be especially powerful with impersonations of influencers, politicians or public figures.
Deepfake ads may be an even more powerful tool to precisely target vulnerable audiences. Today's ad platforms offer unparalleled precision in targeting (spanning demographics, firmographic and intent data) and support rich media formats. Coupled with GenAI's prowess in crafting messages that blend seamlessly into users' feeds, these tools can enable the spread of deepfake content across owned and third-party publishing spaces, maximizing engagement through personalized and native-looking messages.
The numerous elections in 2024 present a ripe opportunity for manipulation. State-sponsored actors can impersonate public figures such as politicians, celebrities and influencers to spread false endorsements, stir civic unrest or influence the outcome of the election.
Breaking The Shackle
Combating deepfakes requires a multifaceted approach that includes regulations, corporate responsibility, public education and anti-deepfake technologies.
• Anti-Deepfake Technology: Corporations, especially major social media platforms, are positioned at the forefront of this battle against the dissemination of deepfake. They should employ sophisticated detection techniques, such as video frame extraction, which have shown promise in identifying AI-generated content by analyzing inconsistencies or anomalies within individual frames of a video. However, detection is just the first step. Upon identifying potentially harmful content, it is imperative to act swiftly to remove such materials to prevent further spread.
• Policies And Regulations: The evolving threat of deepfakes outpaces existing laws, creating a critical need for updated regulations that comprehensively address not just issues related to civic integrity but also privacy, copyright and impersonation. It is imperative for government and regulatory bodies to not only draft and enact new legislation but also ensure its rigorous enforcement. Furthermore, these new legal measures must be dynamic and capable of adapting to the continuous technological advancements in AI to effectively mitigate the risks associated with deepfakes. Ensuring a collaborative effort among lawmakers, tech companies and civil society is crucial in crafting regulations that balance innovation with ethical considerations and public safety.
• Education And Awareness: Initiatives that promote digital literacy and integrity are also essential. By collaborating with educators, non-profit organizations and other stakeholders, social media and tech companies can help foster a more discerning online community capable of critically evaluating the information they encounter. Ultimately, we need to empower individuals to critically evaluate media, recognizing the potential for manipulation and the importance of digital literacy—which is easier said than done.
Conclusion
Deepfakes and their deployment mark a new era of vulnerability, leaving none of us immune to the potential chaos unleashed by state-sponsored or financially motivated actors.
With the ability to launch these sophisticated attacks both easily and economically, they pose a significant threat to all, particularly the young, elderly and those with less technical literacy. We must unite in a comprehensive defense strategy, blending legal, technological and educational efforts to protect digital media integrity and minimize deepfake misuse.
This is a shared responsibility we must all embrace to ensure the safety and trustworthiness of our digital world.
Disclaimer: The opinions and viewpoints presented in this article are solely those of the author, Praveen Gujar, and do not reflect the positions or perspectives of his employer or any affiliated organizations.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
