Slashdot Mirror

← Back to Stories (view on slashdot.org)

Microsoft Search Advertisers Get Personal

Posted by CmdrTaco on from the blurring-the-lines dept.

Alascom writes "According to this AP report, Microsoft is raising privacy concerns by allowing search advertisers to use personal information. Yusuf Mehdi, a corporate vice president with the MSN unit, said Microsoft has gathered this personal information by tracking users who have logged into its Hotmail e-mail program or other Microsoft Web sites in order to allow advertisers to target their ads to a specific audience."

4 of 185 comments (clear)

  1. Nothing new by suso · · Score: 4, Interesting

    Did you really expect anything less from Hotmail or Microsoft? I mean come on, this is a website that asks you (auto selected) if you want to subscribe to about 50+ different personalized newsletters when you sign up. I don't see how people can expect their privacy to be respected when the service is free.

  2. Personalised isnt a bad thing by scenestar · · Score: 5, Interesting

    The biggest problem with advertising reamins that products advertised are not interresting to me.

    I have yet to find a advert for cheap merch pressing or jobs for young webdesigners. Instead i get bombarded with ads for crazy frog ringtones.

    As long as they dont read my email line by line it's FINE with me

    --
    perpetually dwelling in the -1 pits
  3. An international view by Anonymous Coward · · Score: 5, Interesting

    I live in Sweden, where this sort of thing is illegal unless you ask your customer for permission first.

    Could this spell problems for Microsoft Sweden, or MS in any other (most likely european) country with the same laws?

    Yes, MS has registered and is using hotmail.se.

  4. Re:Not really personal information... by chialea · · Score: 4, Interesting

    >Notice they aren't releasing any information (like your name, etc.) that would explicitly identify the person to the advertisers.

    There is some rather interesting research on this. Gender, age, and location is enough to identify most people, since, really, there are a limited number of people in each cartesian-type category. Note that this is different from aggregate data, which may be less personally identifiable, depending on how it's done.

    The people at the data privacy lab have gone through and identified people in "non-personally identifiable" information released by several sources. Part of the problem is that you can put these sources of data together with high confidence and both narrow down individual people and gather a LOT of information about them. I'm sure they have some papers up if you're interested:

    http://privacy.cs.cmu.edu/

    Lea