JPMorgan Chase Builds Tech For YouTube Screening; EU Weighs New Regulations Around Data

Here’s today’s news round-up… Want it by email? Sign up here.


JPMorgan Chase isn’t satisfied with YouTube’s brand safety review, so it’s taking matters into its own hands. The bank’s in-house programmatic team has begun to whitelist channels where it can buy media at scale, Business Insider reports. With 17 filters, the bank uses an algorithm to identify red flags, while also monitoring comments. “The model Google has built to monetize YouTube may work for it, but it doesn’t work for us,” says Aaron Smolick, JPMorgan Chase’s executive director of paid media analytics and optimization. “The attention of protecting a brand has to fall on the actual people within the brand itself.” More.

Data Dealings

European competition chief Margrethe Vestager is taking a magnifying glass to big data. The EU is considering an overhaul of its M&A rules to assess the value of data in addition to revenue and price. “In some areas, these data are extremely valuable,” Vestager tells The Wall Street Journal. “They can give the parties that have them immense business opportunities that are not available to others.” But the difficulty in determining the value of data and its potential scale make it a tough metric to regulate. Forcing companies to share proprietary data could also stifle innovation and allow smaller companies to profit from incumbent companies’ investments, says Nick Wallace, a policy analyst at the Center for Data Innovation. “If policy makers fail to see that distinction, then it could stifle innovation in Europe.” More.

Better Recognize

Image recognition technology has made tremendous strides in the past couple of years, and coupled with advancements in artificial intelligence, it’s unlocked valuable insights buried in large photo sets. “For the first time … researchers are able to analyze large quantities of images, pulling out data that can be sorted and mined,” writes Steve Lohr at The New York Times, covering a recent Stanford research project based on 50 million Google Street View images. By analyzing the cars in driveways, the Stanford researchers were able to accurately predict income, race, education and precinct-level voting across the country. More.

Flip The Script

Pretty much any browser or website with an account login now comes with a password management tool to autofill contact info like an email address or username and password. But a Princeton computer science team identified a couple of software tools, AdThink and OnAudience, that extract data from the login field after it autofills. “That information can then be used as a persistent ID to track users from page to page, a potentially valuable tool in targeting advertising,” reports The Verge. Audience Insights, which operates AdThink, matches the data to Acxiom’s consumer identity set. “We’d like to see publishers exercise better control over third parties on their sites,” says Princeton professor Arvind Narayanan. “These problems arise partly because website operators have been lax in allowing third-party scripts on their sites without understanding the implications.” More. And check out the Princeton report.

But Wait, There’s More!

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!