Pretending to be on the Zoom could get a little trickier, as enterprise browser company Surf Security has a deepfake detector that it says is the real deal.
The vendor’s model makes determinations by analyzing patterns in the voice spectrum, the Surf team told IT Brew, and works with browser-based SaaS audio services like Slack, Zoom, and Microsoft Teams.
“You don’t even know that I’m checking whether you’re fake or not,” CTO Ziv Yankowitz said on a demo on Google Meet—one that confirmed in a few seconds that this IT Brew reporter was human. (Phew!)
The company released a beta version of the push-button feature on November 20.
Fake news! The US Department of the Treasury’s Financial Crimes Enforcement Network (FinCEN) released an advisory in November to warn financial institutions of an increase in fraud schemes using deepfake media, including the creation of false documents, photos, and videos to defeat customer identification processes.
Identity verification company Entrust recently reported an “all-time high” in deepfake threats over the past year: One every five minutes, according to the firm.
Deloitte predicts that GenAI-related fraud could lead to losses of $40 billion in the US by 2027. Some deepfakes have even found their way into the interview process.
Just browsing! Jeff Pollard, VP and principal analyst at Forrester, acknowledges the Surf tool is one of the earliest deepfake detectors directed at an enterprise user, but sees the feature as one that is, by design, limited to the browser.
Top insights for IT pros
From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.
“Let’s say that you’ve got a call center. This browser control from a call center perspective is entirely irrelevant, because the way you’re biometrically authenticating users is either via apps or via voice print, or based on phone calls,” he told IT Brew.
“We should be much more focused on stopping [employees] from opening phishing emails than worrying about someone deploying a deepfake against them, right? It’s the basics that still catch people, not so much this super cool, advanced, fancy stuff,” Pollard said.
Yankowitz said browser-based call-center clients have expressed interest in the technology.
New deepfake detectors have joined the market this year. Security companies like Trend Micro and McAfee have released tools. Social media services, including Meta and TikTok, have enabled features that apply hidden signals, or digital watermarks, to a sample. Some detection options, compared to Surf’s seconds-long, real-time inspection, require an upload of audio and video.
“A lot of people are taking a swing at how to solve this problem, because it is going to be a big problem,” Pollard said.