Deepfakes: when AI hits fake news│Disinformation with Andrea G. Rodríguez
Understanding the Impact of AI on Disinformation
The Rise of Deepfakes and Disinformation Campaigns
- Videos featuring real people making statements they never did are prevalent, marking a new era in disinformation through deepfakes.
- These campaigns exploit societal vulnerabilities to spread confusion, erode trust, and create divisions within communities.
- Artificial intelligence (AI), particularly algorithms that process large datasets, plays a crucial role in creating realistic audiovisual content known as deepfakes.
- The creation of deepfakes involves neural networks that mimic human brain functions, raising concerns about the authenticity of what we see online.
Risks Associated with AI in Disinformation
- Software for generating deepfakes is commercially available; notable examples include scenes from "Star Wars: Rogue One" where characters were artificially created.
- In 2019, researchers used GPT-2 to generate fake news imitating credible sources like The New York Times, demonstrating AI's ability to produce convincing text.
- This technology can enhance the speed and credibility of disinformation; GPT-2 was so effective it had to be reviewed due to potential misuse.
Microtargeting and Its Implications
- Microtargeting strategies, exemplified by Cambridge Analytica's work during elections, analyze personality traits to tailor marketing campaigns effectively.