Deepfakes refer to manipulated media files; mostly videos or audio with real people performing physical activities or speech the subjects never did or said. Deepfakes are created using advanced AI techniques, largely generative adversarial networks capable of generating synthetic content that looks and feels very real. A great deal of concern has been expressed in using this deepfake technology for blackmail, defamation, and political manipulation.
The Impact of Deepfakes
Erosion of Trust
Deepfakes can be a part of the process for the erosion of trust in institutions, persons, and the media. If one cannot actually know what’s real or fake, then everything’s suspect.
Political Manipulation
Deepfakes can be used to spread misinformation and disinformation about political candidates or events, potentially influencing elections and public opinion.
Personal Harm
Deepfakes can be used to target individuals with malicious intent, leading to personal harm, reputational damage, and even legal consequences.
The Role of Generative AI
Deepfakes essentially revolve around generative AI a branch of AI that deals with the creation of new content. Generative models can be trained on large data sets of real world images or videos or even audio, learning these patterns to generate highly realistic synthetic content. In this manner, deepfakes that are very hard to distinguish from real content can be created.
The Disinformation Crisis
Deepfakes represent just the tip of the iceberg in an emerging disinformation crisis in the digital age. The wide reach of social media, combined with the sheer speed at which content may be created and circulated, democratizes the means through which actors hell-bent on mischief can peddle false or misleading information. The resultant effect may be to influence public opinion, erode trust in institutions, and even destabilize societies.
Countering the Threat of Deepfakes
To combat the threat of deepfakes and disinformation, a multi-faceted approach is needed. This includes:
- Technological Solutions: Developing advanced tools and techniques for detecting and identifying deepfakes.
- Unmasking the False: Advanced Tools and Techniques for Deepfake Detection
- Education and Awareness: Raising public awareness about the dangers of deepfakes and how to spot them.
- Regulation and Policy: Implementing laws and regulations to address the misuse of deepfake technology.
- Generative AI and deepfakes: a human rights approach to tackling harmful content www.tandfonline.com
- International Cooperation: Establishing international cooperation to combat the global spread of disinformation
Deepfakes and Generative AI
| Application | Positive Use Cases | Negative Use Cases |
|---|---|---|
| Entertainment | Creating realistic historical reenactments or fictional characters | Spreading misinformation or propaganda |
| Education | Simulating historical events or scientific concepts | Creating fake news or deepfakes to manipulate public opinion |
| Healthcare | Training medical professionals on complex procedures | Creating fake medical advice or diagnoses |
| Research | Generating synthetic data for research purposes | Creating deepfakes to blackmail or harass individuals |
| Art and Design | Creating unique and original artwork | Infringing on copyright or intellectual property |
| Product Development | Testing products in virtual environments | Creating fake product reviews or endorsements |
Real-World Examples of Deepfakes and Generative AI
- Historical reenactments: Deepfakes have been used to recreate historical events, such as the assassination of Julius Caesar, in a visually stunning and informative way.
- Fictional characters: Deepfakes have been used to bring fictional characters to life, such as in the case of the “Star Wars: The Clone Wars” series, where deepfakes were used to create a younger version of the character Obi-Wan Kenobi.
- Language learning: Deepfakes have been used to create realistic language learning scenarios, allowing students to practice their language skills in a more immersive environment.
- Medical training: Deepfakes have been used to create simulated patients for medical students to practice on, helping them to develop their clinical skills.
- Testing products: Deepfakes can be used to test products in virtual environments, reducing the need for physical prototypes and saving time and money.
- Customer feedback: Deepfakes can be used to gather customer feedback on products, allowing companies to identify areas for improvement.
Deepfakes have been used to spread misinformation and disinformation, particularly in the political arena. For example, deepfakes of politicians making false or misleading statements have been used to manipulate public opinion.
Deepfakes have been used to blackmail individuals or organizations, threatening to release embarrassing or damaging content if demands are not met.
Deepfakes have been used to harass and bully individuals, particularly women and minorities.
Ad campaigns used deepfake and generative AI
Nike’s “Forbidden” Ad
This ad, featuring a deepfake version of Ronaldo, made quite a stir when it came out. They created an impressively real version of the soccer star who went on to nail an amazing goal. What it was doing was showing that the best athlete also has difficult times, but it is making a statement on how realistic this technology can seem
Samsung’s “The Future” Ad
This ad from Samsung features a deepfake version of Dr. Serkan Toto, a well-known futurist. In the ad, Dr. Toto talks about the future of technology, and how Samsung is at the forefront of it all. The ad does a great job of showcasing the potential of deepfake technology, while also highlighting Samsung’s products and services.
#whereeveryouare campaign
Deepfakes of Cara Delevingne were used by the fashion brand Zalando in its #whereeveryouare campaign. Provided with deepfake technology – and footage of Delevingne – the brand could create as many as 290,000 localised ads for towns and villages across Europe. None of this would have required sitting for hours in one location, learning several languages, practicing pronunciation, or running through several takes. Although the idea is that influencers are going to jack up their prices when their image is used this way, deepfake technology actually reduces the time, money, and effort invested in producing an ad campaign. The fashion and beauty industries have also pointed to other possible use cases, citing how deepfake technology can be used to show products on models with different skin tones, heights, and weights to display new collections to consumers.
Softlogic Life’s commercial in Sri Lanka
It is the first-ever deepfake video campaign in the insurance industry, the fastest growing life insurer in Sri Lanka, Softlogic Life Insurance PLC. In taking this revolutionary step, they brought back four legends of music: H.R. Jothipala, Milton Mallawarachchi, Clarence Wijewardena, and Sunil Perera, to convey a powerful message as to why life insurance is necessary more than ever, especially in these uncertain times.
These are just a few examples of how deepfakes and generative AI are being used in the real world. As technology continues to advance, we can expect to see even more innovative and potentially harmful applications of these powerful tools.
Drop a comment with your thoughts on the Deepfake and Generative AI.