• by Wendy Davis  @wendyndavis, March 2, 2012

The ad industry's self-regulatory group Digital Advertising Alliance recently drew praise from some advocates for promising to honor a browser-based do-not-track header. But the news also renewed questions about what "do-not-track" means to ad companies.

For many years, the industry has taken the position that companies should not serve targeted ads to people who opt out of behavioral advertising, or collect information for the purpose of ad targeting.

The Digital Advertising Alliance also prohibits members from collecting any information about Web users in order to determine their eligibility for employment, credit, health care or insurance. But self-regulatory principles allow companies to track consumers for purposes like analytics, frequency capping and site optimization, even when consumers have opted out of online behavioral targeting.

But some advocates say that companies should completely stop gathering data about consumers when they opt out of online ad targeting.

On Friday, Federal Trade Commissioner Julie Brill appeared to agree with those advocates. "For me, one of the most critical points is that Do Not Track is not just Do Not Target ... but also, when the consumer so chooses, Do Not Collect," she said in a keynote address at a privacy conference held at Fordham Law School.

She added that while she was pleased with the DAA's stance on browser-based do-not-track tools, she intends to "closely watch industry’s progress in creating a robust solution to effectuating consumers’ choices, including whether choices provided to consumers address the collection of their information in the first instance, and not just the receipt of targeted ads."

Brill did not elaborate on whether she envisions exceptions for analytics, fraud detection or other purposes.

The commissioner also questioned whether "heightened protections" for consumers are needed when companies used relatively mundane data to draw conclusions about sensitive conditions. She referenced recent newspaper articles about how Target deduced certain consumers' pregnancies, and said the practice raised concerns even if the predictions were based on "innocuous data," like the purchase of lotions and newborn-size diapers.

"The same type of innocuous data could be used to make other predictions of a sensitive nature, like sexual orientation, financial status, and the like," she said. "We need to address whether heightened protections should be required in this type of situation. ... At first blush, it seems that some form of heightened protections are in order."