Hot topic
Google’s head of privacy is departing in the fall and will not be replaced, news that broke just days after a leak revealed the company had inadvertently committed various privacy violations over a number of years, including collecting childrens’ voice data.
Keith Enright posted on LinkedIn on Wednesday that he is “ready for a change” after 13 years at Google. His departure was first reported by Forbes, which stated that neither him nor Matthew Bye, Google’s head of competition law, will be replaced when they leave.
Enright’s post was met with an outpouring of well wishes and support from dozens of privacy professionals; there was no mention of the leak and Google’s violations over the years.
The leak details thousands of privacy violations that happened between 2013 to 2018. Documents detailing the incidents were shared with 404 Media.
In one incident, a Google speech service logged audio of an estimated 1,000 children for about an hour, the report said. Another involved the disclosure of more than a million email addresses of Socratic users, a mobile learning app that Google acquired in 2018. Meanwhile, YouTube was found to be recommending videos based on users’ deleted watch histories. Google told 404 Media that every incident “was reviewed and resolved at that time.”
Google has shelled out millions for privacy violations over the years. In February, it agreed to a $350 million settlement related to a security breach at its Google+ social media website, which exposed user data for three years. Google+ was shut down in 2019. In April, the tech giant agreed to destroy the private browsing history of millions of people in the U.S. as a part of a settlement related to how it tracks “incognito” users in its Chrome browser.
Under the hood
The advertising industry is bracing for a brand safety crisis as ads are introduced into experimental AI products like Google’s AI-powered summarized responses, which has faced consumer backlash for surfacing conspiracy theories and recommending dangerous acts.
Google said last week it blocked some results and implemented “more than a dozen technical improvements to our systems” after its AI Overviews were discovered to be producing misleading responses following the product’s launch in mid-May.
Technology and advertising experts I spoke with blamed the hastiness with which AI products are launching for their predisposition for errors. This is the third time Google has come under heat this year for issues within its AI products; executives urged the company to conduct more controlled testing to work out kinks before using the public as “guinea pigs.”
But the search giant is not alone in speeding up product development in an attempt to win share in the AI race; Amazon, Meta and Microsoft have also significantly ramped up their computing resources to fuel their AI ambitions. OpenAI CEO Sam Altman was fired in November over allegedly clashing with the board about how quickly the company should develop AI tools. He returned to the role a few weeks later.
“In Google’s defense, we are seeing this pretty much everywhere in the industry right now,” said Valerie Carlson, chief creative officer of Critical Mass. “There is so much fear that you will not be first, that you will not be fast enough. There is not the right balance between the human touch and the machine.”
With ads “soon” coming to the AI overviews, media buyers have concerns about potential brand safety issues, made worse by their inability to control placements. Advertisers who purchase search, shopping or Performance Max campaigns cannot opt out of appearing within the AI overviews.
A media executive shared on background that it feels like the ad industry is waiting for a bad placement to occur against an AI response “for everything to topple back.”
Fresh tech
- Nvidia’s market capitalization surpassed $3 trillion on Wednesday, making it the second-most valuable company in the world, behind Microsoft. The chipmaker, which is profiting from the AI boom, has outpaced the likes of Apple, Amazon and Alphabet.
- Ticketmaster owner Live Nation revealed on Friday it was investigating a data breach after a group of hackers called ShinyHunters claimed they had stolen the personal details of 560 million customers. The hackers claim the stolen data includes names, addresses, phone numbers and partial credit card details, and are reportedly demanding a $500,000 ransom.
- Swedish telco Ericsson plans to shutter its ad business, Emodo, after it failed to gain momentum, Business Insider reports. Emodo launched in 2017 as a data management platform and acquired demand-side platform Placecast in 2018 and mobile ad exchange Axonix in 2021.
- Humane has alerted owners of its AI Pin to “immediately” stop using the charging case because it “may pose a fire safety risk,” The Verge reports.
- Scope3 has launched a reporting tool for brands and media buyers that ingests campaign data to measure the ad emissions of digital ad campaigns and compare it to market benchmarks.
- Marketing services firm Quad has installed the Covision Media 3D-Twin scanner at its studio in North Carolina to enable the creation of “digital twins” for clients.
- LinkedIn is testing in-stream video sponsorships with Bloomberg, Business Insider and Dow Jones. It has also expanded the AI capabilities of its automated B2B campaign creator, Accelerate, including integrating Microsoft’s AI-powered graphic designer into the tool.
- X’s de facto spokesperson Joe Benarroch is reportedly leaving the social media platform. He joined last June following Linda Yaccarino’s appointment as CEO. The two worked closely together at NBCUniversal.
- Elsewhere, X has updated its content policies to officially permit adult content, including nudity and pornography, provided it is “consensually produced and distributed” and properly labeled. The policy clashes with the positions of the largest social media platforms, which either heavily restrict or entirely ban adult content.
Regulatory developments
The Justice Department (DOJ) and the Federal Trade Commission (FTC) have reached a deal that will enable them to conduct antitrust investigations into Microsoft, OpenAI and Nvidia’s outsized impact on the AI industry. Under the deal, the DOJ will take the lead in investigating Nvidia for potential antitrust violations, while the FTC will take on OpenAI and Microsoft. The New York Times revealed the details of the deal, expected to be finalized in the coming days.
Temu has been designated as a “very large online platform” under the European Commission’s Digital Services Act (DSA) after meeting the regulator’s threshold of servicing at least 45 million active monthly users in the E.U. It means Temu will have to comply with the most stringent rules under the DSA by the end of September, including a requirement to conduct risk assessment reports related to how illegal products are sold on its e-commerce platform and put solutions in place. It will also have to publish repositories of all the ads served on its platform.
New York state plans to ban social media companies from serving children with algorithmically-delivered feeds without parental consent, The Wall Street Journal reports. The bill, expected to be voted on this week, would also prohibit platforms from sending notifications to minors late at night without parental consent.
Google issued guidance to U.S. advertisers last week detailing how it will comply with five privacy law provisions from Florida, Texas, Oregon, Montana and Colorado being enforced this year. The new laws, including a universal opt-out mechanism in Colorado, are designed to restrict the data that businesses can use and empower users to control their data. This will impact advertisers’ ability to target ads and measure performance in those states.
Clashes and debates
The Trade Desk’s list of top 100 publishers has caused a furor in the industry over what designates a “premium” publisher and why diverse and local news outlets are not better featured. Sparrow Digital’s principal and cofounder Maja Milicevic argued in a LinkedIn post the list “opens up more questions than answers.”
Buddy up
- Apple has reportedly struck a deal with OpenAI to integrate its products into iOS, according to The Information. The iPhone maker is expected to announce a slew of AI improvements, including to its Siri voice assistant, at its annual Worldwide Developers Conference next week. Bloomberg reported that Apple was planning to integrate ChatGPT into the new iOS 18 upgrade.
- Customer experience management platform Sprinklr is utilizing Reddit’s Data and Advertising APIs to allow clients to analyze public conversations on Reddit to inform their paid and organic strategies on the platform.
Trends
Investigative nonprofit Global Witness recently submitted election disinformation related to the upcoming European parliamentary elections in Ireland to TikTok, YouTube and X to evaluate whether the ads would be flagged by the tech companies’ systems.
The ads, submitted in May, included false claims about how to vote in the elections, including urging users to text a fake number instead of voting in person.
It said TikTok approved all 16 of the video ads to run, YouTube approved two out of 16 ads and X approved none of them. X also suspended the account which submitted the ads “due to a policy violation,” the nonprofit said.
Global Witness also tested whether TikTok would flag a political ad, which it does not allow, once it goes live. It said an ad that read “It’s an election year!” garnered over 12,000 impressions in less than an hour.
The nonprofit said it will share the findings with the EU Commission as “evidence of TikTok’s failure” to comply with the DSA.
Should brands be wary of advertising on TikTok amid rising political disinformation?
Dollars and deals
Contextual ad firm Seedtag acquired sell-side platform Beachfront to bolster its connected TV offering.
Reading list
Chinese firms are rebranding and creating U.S.-domiciled businesses in an attempt to protect their operations in the U.S. as they face a growing crackdown from the Biden administration, the Journal’s Heather Somerville reports.