Facebook will offer a cash prize to the person or group that can build a next-generation way to detect deepfakes — and it’s even creating its own original deepfakes for participants to work with.
The Deepfake Detection Challenge is a partnership between Facebook, Microsoft, the Partnership on AI, and academics. Facebook has put up $10 million towards the partnership to help detect and deal with videos and other media that have been manipulated to show someone doing or saying something they never said or did.
“The goal of the challenge is to produce technology that everyone can use to better detect when AI has been used to alter a video in order to mislead the viewer,” wrote Mike Schroepfer, the Chief Technology Officer of Facebook AI, in a blog post announcing the program.
The challenge will use a brand new set of videos that feature paid actors, so no Facebook user data will be part of the program. The company will create both unmodified and “tampered” videos using different deepfake AI techniques for participants to detect. The hope is to develop a way to better detect and prevent tampered media — and give social network like Facebook a way to flag deepfakes before they go viral.
To encourage participation, there will be research collaborations and prizes as part of the challenge. Facebook has already dedicated $10 million is being dedicated to the project.
“[The] $10 million total is being given across university grants, challenge awards, workshops, and starting to build the dataset. We will continue to invest to further the development of the dataset over time and support the research community,” a Facebook spokesperson told Digital Trends.
The Deepfake Detection Challenge will begin in October and run through May 2020 globally.
Deepfakes are a relatively new technology that has grown rapidly this year. There’s already been issues with fake deepfakes purporting to show famous people saying something they never actually said, including one deepfake of Facebook CEO Mark Zuckerberg proclaiming his power over “millions of people’s stolen data.” Experts say that the technology is advancing quickly and deepfakes will only get more convincing (and easier to create) over the next few years.
Experts say that to fight the perpetuation of deepfakes being used in the wrong way, programs like the Deepfake Detection Challenge are essential.
“Technology to manipulate images is advancing faster than our ability to tell what’s real from what’s been faked. A problem as big as this won’t be solved by one person alone,” said Phillip Isola, Assistant Professor of Electrical Engineering & Computer Science at the Massachusetts Institute of Technology, in the announcement. “Open competitions like this one spur innovation by focusing the world’s collective brainpower on a seemingly impossible goal.”
AI companies spend enormous amounts of time convincing us their technology is safe and secure. But what happens when those same companies become targets of increasingly sophisticated AI-powered attacks? Anthropic, the company behind Claude, is now facing that question. According to a Washington Post report, cybersecurity startup Cathedral discovered vulnerabilities that let its researchers access sensitive information in Anthropic's internal systems.
The researchers used AI agents to identify potential entry points in less than 12 hours. Interestingly, they didn't need to crack Claude itself. Instead, they exploited weaknesses in third-party business applications that Anthropic uses for everyday operations. Anthropic's security testing program disclosed the findings, which allows outside researchers to identify and report vulnerabilities.
Military working dogs charge into places the rest of us would flee. So, I understand why anyone would want to breed tougher, healthier dogs for the job.
Warhound, a new startup from Colossal Biosciences co-founder Ben Lamm and Anduril co-founder Palmer Luckey, has stepped out of stealth with $200 million to clone and gene-edit working dogs for American troops, police departments, and emergency services (via TechCrunch).
AI can describe a picture, identify objects, and answer questions about what a camera sees. But when a blind user needs a quick answer, getting an entire spoken essay about the scene isn't exactly helpful. Researchers at the University of Maryland (UMD) have found that AI-powered visual assistance can leave blind users doing too much work to get the information they actually need.
The study, titled Say It My Way: Exploring Control in Conversational Visual Question Answering with Blind Users, examined how people interact with Be My AI, an AI-powered feature of the Be My Eyes app. The researchers found that users often received lengthy descriptions that didn't adequately address their immediate goals, forcing them to ask follow-up questions or find ways to steer the AI towards a more useful answer.