”My Deepfake Lied About That,” and Meet the Oppenheimer of AI

“Deepfake” of the Week

Tucked within the court filings of a lawsuit against Tesla (for an alleged death caused by the company’s Autopilot technology) is a high-profile legal first: blaming deepfakes for public statements made by CEO Elon Musk. 

In this case, Tesla’s legal team stated, “[Musk], like many public figures, is the subject of many ‘deepfake’ videos and audio recordings that purport to show him saying and doing things he never actually said or did.”

One of the alleged deepfakes in question is a video of Musk being interviewed in 2016 during the Recode conference, where he made claims on how then-current Tesla models could drive better than people. Deepfake technology barely existed in its infancy around this time, with the term “deepfake'' first appearing the following year. Thus, the probability of these videos being deepfakes, especially as they were recorded and streamed live by Recode, hovers around zero.

That said, we ran the video through our platform.

What does Reality Defender's platform say?

What did the presiding judge say?

“Their position is that because Mr. Musk is famous and might be more of a target for deep fakes, his public statements are immune. In other words, Mr. Musk, and others in his position, can simply say whatever they like in the public domain, then hide behind the potential for their recorded statements being a deep fake to avoid taking ownership of what they did actually say and do. The Court is unwilling to set such a precedent by condoning Tesla's approach here."

Geoffrey Hinton Leaves Google to Sound Alarm Bells on AI

Cognitive scientist Geoffry Hinton, known for his extensive work and pioneering in artificial neural networks, has left his ten-year stint at Google to speak freely on the dangers of widespread use of AI — including the use of technology he developed. Concerned about everything from bad actors weaponizing AI to catastrophic job loss, Hinton spoke at length with The New York Times on what he fears in the near future from rampant use and deployment of advanced artificial intelligence-driven technologies.

The First Major American Political Ad to Feature Generative Content

Shortly after President Joe Biden announced his reelection campaign, the Republican National Committee released an ad featuring generative video content to demonstrate what they envisioned a post-Biden America might look like. This marks the first major use of generative video content in an American political ad on a national level.

TikTok’s AI Avatars

Millions of TikTok users are about to gain access to the company’s AI avatar models, according to a new leak. Similar to the app-based AI avatar craze from earlier in the year, users will be able to stylize themselves (or whatever photo they upload) using “generative AI technology.” TikTok is currently facing bipartisan calls to ban the app from U.S. soil, which might make this new avatar generator (and everything else) a tad short-lived.

How a Tier-One Bank Stopped Deepfaked Audio with Reality Defender

Facing an increase in sophisticated incidents utilizing deepfaked audio to deceive employees, a multinational tier-one bank contacted Reality Defender to analyze calls and determine real callers from those attempting to commit fraud.

See how Reality Defender met our clients’ bespoke needs in fighting financial fraud.

Download Case Study

Reality Defender on CNBC

Reality Defender Co-Founder and CEO Ben Colman appeared on CNBC this past Friday to discuss deepfake detection with Last Call host Brian Sullivan.

See how hard it was to spot deepfakes without comprehensive detection in the clip here.

Deepfake News

• An Andy Warhol copyright case in the U.S. Supreme Court could have a material impact on the future of generative content. (Wired)
• A Wall Street Journal reporter “cloned” themselves with AI and fooled everyone — including their family. (The Wall Street Journal)
• The new Indiana Jones film features a lengthy sequence with a de-aged deepfake replica of Harrison Ford. (Deadline)
• A Canadian man is in prison after he was caught generating deepfaked CSAM. (New York Post)
• More on the fallout from the RNC deepfaked political ad. (Vice)

AI News

• Amazon is working on their own LLM to improve Alexa. (TechCrunch)
• Meanwhile, Apple allegedly faces internal turmoil due to the slow AI development of Siri. (The Information)
• Generative AI could upend the U.S. patent system. (Financial Times)
• A proposed law hopes to prevent AI from launching nukes. (Ars Technica)
• Meta looks to bring “AI agents” to its billions of users. (The Verge)
• The EU has proposed new copyright rules for generative AI. (Reuters)