Deepfake

Timeline

• 1978The artist worked with Deepfake technology to create an AI actor, a synthetic version of 80s movie star Ornella Muti, travelling in time from 1978 to 2018.
• 1990Deepfake technology has been developed by researchers at academic institutions beginning in the 1990s, and later by amateurs in online communities.
• 2018Commercial development In January 2018, a proprietary desktop application called "FakeApp" was launched.
• 2019As of 2019, "FakeApp" had been largely replaced by open-source alternatives such as "Faceswap", command line-based "DeepFaceLab", and web-based apps such as DeepfakesWeb.
• 2020A survey of deepfakes, published in May 2020, provides a timeline of how the creation and detection of deepfakes have advanced over the last few years.

Read

• Trouble with DeepfakesJosh Gregory · 2024Book
• DeepfakesLoveleen Gaur · 2022Book
• Deepfakes and Their Impact on BusinessGaurav Gupta · 2024Book
• Deep FakeAlex Schuler · 2021Book

Listen

• Understood: Artificial IntimacyCBCPodcast
• DeepfakeUnknownPodcast
• The Deepfake DialoguesAshmita RajmohanPodcast
• Deep FakePodshapePodcast

By the numbers

• 73.6Internet users (%) — global, 2025 (World Bank)
• 111.5Mobile subscriptions / 100 — global, 2025 (World Bank)

Debates

• Should the creation of deepfakes be legally restricted?One view: Yes, to prevent the spread of misinformation, fraud, and reputational damage. · Another: No, such restrictions could stifle artistic expression, satire, and legitimate research.Open question
• Are social media platforms responsible for detecting and removing deepfakes?One view: Yes, platforms have a duty to moderate harmful content and protect their users from manipulation. · Another: No, it's technically challenging to detect all deepfakes, and platforms should not be arbiters of truth.Open question

Glossary

• DeepfakeSynthetic media, typically video or audio, created using artificial intelligence to manipulate or generate content.
• Artificial Intelligence (AI)Computer systems designed to perform tasks that usually require human intelligence, such as learning and problem-solving.
• Machine LearningA subset of AI that enables systems to learn from data, identify patterns, and make decisions with minimal human intervention.
• Generative Adversarial Network (GAN)A type of AI algorithm used to create new, realistic data instances, often employed in deepfake generation.
• Synthetic MediaAny media content, including images, audio, or video, that is artificially generated or altered by AI.
• MisinformationFalse or inaccurate information spread, regardless of intent to deceive.

Careers

Roles this can lead toward

Student research

Published policy papers by One Young India delegates — every delegate leaves published under their own name.

• Separating Real and AI Media – Difficulties and SolutionsK.V. Siva Surya
• Deepfake Sexual Exploitation in IndiaAnahat Kaur
• The Public Persona Trap: US College Admissions and Social MediaGunika Chaudhary
• Understanding misinformation in India: The case for a meaningful regulatory approach for social media platformsGandharv Dhruv Madan
• The Evolution of Social Media Marketing: From Engagement to CommerceHimanshi Bhutoria

Threads 4

Where this connects to other fields — and why it's worth knowing.

• Cryptography & Digital Security Mathematics

  When deepfakes make any video look fake-able, staring harder won't save us. We'll trust images the same way we trust money moving between banks: with a cryptographic signature, a secret math stamp that proves a photo is real, like verifying a bank transfer.

• Procedure & Evidence Law

  Deepfakes create a nasty bonus for liars: once any video could be faked, a guilty person can wave away real footage as 'fake'. So even genuine evidence starts losing its power in court. Truth gets easier to deny just because lies got easier to make.

• Evolution Science

  Every time a detector learns to spot fake videos, that lesson is used to build sneakier fakes that slip past it. Predator and prey keep upgrading against each other, exactly like germs and your immune system locked in an endless arms race.

• Money, Banking & Credit Economics

  A banknote and a photo are both worth trusting only because faking them is hard and expensive. Deepfakes make faking video dirt cheap. When counterfeiting gets easy, trust loses value — like money in a crash, reality itself gets "inflated."