DeepTruth: The First Comprehensive Solution to Combat Deep Fakes
Technology has unleashed a new, terrifying prospect on the world: We can no longer believe what we see or trust what we hear. It’s an unprecedented, almost unimaginable threat to our everyday lives and stability. To understand the magnitude of the problem and the urgent need for the solution, we’ve delved into the reasoning behind DeepTruth, a comprehensive solution to combat deep fakes at their source.
The threat. New waves of highly disruptive technologies that merge artificial intelligence (AI) and video-audio creation are entering the marketplace. These technologies — known as deep fakes — facilitate the ability to construct recordings of events and conversations that never actually happened.
These deep fakes use our own words against us: A 15-minute recording is all a deep fake creator needs to compile a library of every sound in the speaker’s native language. Armed with that audio file, a creator can use deep fake technologies to make the speaker “say” anything — new words, sentences, phrases, conversations, speeches. And with the speed of viral media, these recordings could easily be distributed around the globe in 24 hours or less.
The urgency. Anyone, anywhere, can weaponize fake information. If unchecked, the consequences will be devastating on both a personal and global scale. If the average person cannot tell the difference between faked and real media, deep fakes will cause consumers to suspect all media. The ramifications will reverberate both personally and globally: Deep fakes will start wars, destroy careers, skew elections, ruin businesses, manipulate stock markets, shatter relationships, devastate reputations, convict the innocent, exonerate the guilty.
Why the current approach won’t work. The prevailing strategy to manage this crisis is purely defensive. Social media platforms, government, and the media have taken an investigative approach instead of tackling deep fakes at the source. This approach is bound to fail because architects of deep fakes will continually improve their technology to stay one step ahead of those trying to identify and expose their counterfeit productions. By concentrating on identifying deep fakes after their creation, seekers of the truth will always be one step behind.
Further complicating matters is the increasing velocity and volume of deep fakes. It will simply take too long for anyone to identify, examine, publish, and distribute responses to fake creations. Companies and everyday people will not have the resources or technology to identify them in time. The viral nature of our connected world — and the ever-increasing speed at which society passes judgment — will lead to irreversible consequences. False information will succeed, time after time.
Why DeepTruth is the antidote to deep fakes. The current approach to fighting deep fakes places the burden of truth on media consumers rather than producers. It’s a passive strategy that perpetually gives the advantage to makers of deep fakes. Exhausted consumers will eventually grow weary of researching the truth and lose trust in media.
The solution is to establish media as immutable and irrefutable during its creation. This approach eliminates the burden of truth for consumers and instead shifts it to the authors of video recordings, audio recordings, or photographs.
How it works. DeepTruth uses a variety of technologies to verify and validate media during creation. The platform authenticates, in real time, layers of metadata that are inaccessible to content producers and distributes that metadata to a public ledger. This metadata creates a digital fingerprint, one that irrefutably proves media is authentic and unedited.
This metadata includes, but will not be limited to:
The metadata and the encryptions, or hashes, of the data will be stored separately from the recorded media through distributed public ledger systems using blockchain technologies. This will ensure an immutable, undeniable endorsement of all data.
- Rotary encryption or hashing
- User data
- Device metadata
- Biometric data
- Remote system data
- Detached time-sensitive data
Countering deep fakes with DeepTruth. Any person will have the ability to verify and distribute any validated recorded media (or subset of that media) created with DeepTruth. A DeepTruth user will be able to instantly identify changes to the original media.
Most importantly, the ease and speed in which people will be able to identify fakes will be immediate. Users will be able to share or download media and evaluate it themselves. Ultimately, a hosted solution (such as YouTube Verified or Vimeo Verified) will be able to do that work for them.
DeepTruth restores trust in media, whether it is generated by a user or a news organization. Its power lies in its undeniable veracity; for once, any person will be able to believe what he or she believes.
It’s proof of the truth.