Deep Learning In Action: Deepfakes

Deepfakes are synthetic media—videos, images, or audio—created using deep learning methods, particularly Generative Adversarial Networks (GANs). GANs involve two competing neural networks: a generator that produces synthetic data, and a discriminator that evaluates its authenticity. Through iterative training, the system learns to produce highly convincing fake content (Goodfellow et al., 2014). This technology enables the replication of facial expressions, speech, and body movements of real individuals, creating media that can be difficult to distinguish from reality.

While deepfakes have potential in fields such as film production, virtual education, and assistive technologies, they also raise serious ethical and societal challenges.

Misinformation and Trust: Deepfakes can be used to fabricate statements or actions by public figures, potentially manipulating public opinion and spreading disinformation (Chesney & Citron, 2019).
Privacy and Consent: Individuals may be featured in deepfaked content without their knowledge or permission, often in harmful or exploitative contexts (Kietzmann et al., 2020).
Regulatory Gaps: Current laws often lack the clarity or scope to address issues of authorship, consent, and liability associated with synthetic media (West, 2021).
Liar’s Dividend: The existence of deepfakes may allow real events to be dismissed as fake, undermining public trust in video evidence and complicating accountability (Chesney & Citron, 2019).

Deepfakes highlight the dual-use nature of AI technologies. While capable of innovation and social benefit, they also demand strong ethical guidelines, regulatory frameworks, and public awareness to mitigate harm.

References

Chesney, R. & Citron, D.K. (2019) ‘Deepfakes and the new disinformation war: The coming age of post-truth geopolitics’, Foreign Affairs, 98(1), pp. 147–155.

Goodfellow, I. et al. (2014) ‘Generative adversarial nets’, in Advances in Neural Information Processing Systems, 27. Available at: https://papers.nips.cc/paper_files/paper/2014/file/5ca3e9b122f61f8f06494c97b1afccf3-Paper.pdf (Accessed: 18 April 2025).

Kietzmann, J., Lee, L.W., McCarthy, I.P. & Kietzmann, T.C. (2020) ‘Deepfakes: Trick or treat?’, Business Horizons, 63(2), pp. 135–146. https://doi.org/10.1016/j.bushor.2019.11.006

West, D.M. (2021) ‘How to combat fake news and disinformation’, Brookings Institution. Available at: https://www.brookings.edu/articles/how-to-combat-fake-news-and-disinformation/ (Accessed: 18 April 2025).