Earlier this year, Zoom’s terms of service made no mention of AI, but an updated version that took effect in July appeared to give the company broad rights to use “Customer Content” for “product and service development,” including for machine learning and AI.
The tweak sparked backlash after a developer flagged the new wording in an article Monday, prompting Zoom Chief Product Officer Smita Hashim to issue a blog post stating that the company would not use customer videos or chats to train its AI models without user consent.
The company also updated its policies Monday to say that “notwithstanding” the uses outlined in its rules, it “will not use audio, video or chat Customer Content” to train AI without consent — which could still leave the door open to significant data collection.
Privacy advocates bashed the update, with some calling on federal enforcers to investigate whether Zoom’s actions violate its privacy pact with the government.
John Davisson, director of litigation at the Electronic Privacy Information Center advocacy group, called it “an alarming, poorly explained, and likely illegal change by Zoom.”
“Zoom now claims that it has virtually limitless authority to reuse your data for any purpose, including training AI,” he said in an email to The Technology 202. “That’s shocking.”
Rep. Jan Schakowsky (D-Ill.), a top lawmaker in data privacy talks on Capitol Hill, called it “yet another example of a tech company taking away an individuals’ right to their own data.”
“The AI boom has only exacerbated this trend as companies vacuum up any data they can to monetize using AI models,” she added in a statement.
Schakowsky and other congressional Democrats in 2020 called on the FTC to investigate Zoom’s privacy practices, as users flocked to its virtual conferencing products during the early days of the coronavirus pandemic. Lawmakers accused the company of overstating the security of its encryption practices and misleading consumers.
In 2021, the FTC struck a settlement with Zoom requiring the company to implement new security measures and to vet product updates for security flaws before rolling them out. The order also barred Zoom from misrepresenting its data collection practices.
Zoom’s latest AI-related changes may run afoul of those restrictions, Davisson said.
“Zoom is promising control over your data that the company is failing to provide in its new features and terms of service,” he said. “That appears to be a major violation, and it’s something the FTC needs to take a close look at.”
“Zoom has a poor track record of protecting consumers’ data and living up to its promises — as their consent order and 2021 settlement prove,” Schakowsky said.
The FTC declined to comment.
In the blog post, Hashim said Zoom made the changes “to be more transparent about how we use and who owns the various forms of content across our platform.”
“We remain committed to transparency, and our aim is to provide you with the tools you need to make informed decisions about your Zoom account,” she said. “We value your privacy and are continuously working to enhance our services while respecting your rights and preferences.”
The tweaks could also attract the attention of state enforcers, who have previously investigated the company’s privacy practices. Since, more states have passed privacy laws creating new requirements that companies allow users to opt-out of some forms of data collection.
Amazon to meet with FTC in final bid to prevent antitrust lawsuit
Amazon will meet with Federal Trade Commission Chair Lina Khan and fellow commissioners next week in a last-ditch effort to prevent an antitrust lawsuit, Leah Nylen reports for Bloomberg News, citing people familiar with the matter.
Nylen writes: “The FTC, which has been investigating Amazon since 2019 over its online marketplace for third-party sellers, is finalizing a suit against the online retail giant. A meeting between the company and the FTC’s commissioners is often one of the last steps before either suit is filed or a settlement is reached.”
The agency has probed the e-commerce giant for several cases of possible anticompetitive conduct. In June, it sued the company for allegedly tricking millions of customers into enrolling in its Prime service.
Amazon and FTC spokespeople declined to comment to Bloomberg. The company has previously argued that Khan should recuse herself from the case in part due to a paper the agency chair authored as a law student which used the company’s alleged market dominance as a reason to rethink antitrust laws, Nylen notes.
Amazon founder Jeff Bezos owns The Washington Post. Interim CEO Patty Stonesifer sits on Amazon’s board.
AI models undergo White House-backed coherence tests at hacker convention
Leading AI firms including Google, OpenAI, Anthropic and Stability AI have volunteered their latest chatbots and image generators to be tested by hackers to see whether the models can be erred toward malicious behaviors, our colleague Will Oremus reports.
The tests will be the first public “red teaming” contest for AI language models — where hackers try to find ways to make AI systems act awry so their makers can try to fix them before they cause harm — and will happen later this week at the DEF CON hacking convention in Las Vegas.
The Generative Red Team Challenge event “has drawn backing from the White House as part of its push to promote ‘responsible innovation’ in AI, an emerging technology that has touched off an explosion of hype, investment — and fear,” Will writes.
The report adds: “The contest underscores the growing interest, especially among tech critics and government regulators, in applying red-teaming exercises — a long-standing practice in the tech industry — to cutting-edge AI systems like OpenAI’s ChatGPT language model. The thinking is that these so-called ‘generative’ AI systems are so opaque in their workings, and so wide-ranging in their potential applications, that they’re likely to be exploited in surprising ways.”
Schumer pushing for New York to reap benefits of microchips funding
Senate Majority Leader Charles E. Schumer (D-N.Y.) is aggressively pushing for key semiconductor funding to be directed toward his home state of New York, Ana Swanson reports for the New York Times.
The Chips and Science Act signed into law last year devotes billions of dollars in funding to spur domestic semiconductor manufacturing and research. Through the bill, the United States “is now preparing to invest tens of billions of dollars in the U.S. semiconductor industry in an effort to boost chip manufacturing across the country and lessen U.S. reliance on foreign factories,” Swanson writes.
“Mr. Schumer, a longtime China critic, primarily views the investments as critical to reducing America’s reliance on Beijing for a technology that powers everything from cars and dishwashers to missiles and fighter jets,” the report adds.
“But he also saw the opportunity to fulfill a more personal goal: securing investment that could revive the factory towns of his home state, which had been hollowed out through decades of competition with China,” Swanson writes. If the New York investments come to fruition, they could “augment his local political support, attract donations from chip companies to fill Democratic coffers and cement his legacy as a proponent of upstate New York,” the Times report adds.
- Tech trade group TechNet has promoted Meghan Dorn to federal policy director.
- Georgetown Law convenes a three-day event on current tech policy issues beginning today at 9:30 a.m.
- The Aspen Institute holds a discussion on app-based gig work at 2 p.m.
- The Federal Election Commission considers a petition to clarify AI usage in election campaigns on Thursday at 10:30 a.m.
- FTC Chair Lina Khan and Justice Department Assistant Attorney General Jonathan Kanter speak at an American Economic Liberties Project discussion on draft merger guidelines on Thursday at 2 p.m.
