Thursday, August 05, 2021

IPSC Panel 12 – Identity, Data, and Privacy

Dustin Marlan, The Dystopian Right of Publicity

Privacy problems (surveillance) are often analogized to the dystopia of 1984; ROP problems stemming from infinite transferablility can be analogized to Brave New World (1932). A state of unfreedom that is apparently chosen and pleasurable (though enforced by drugs and conditioning). This is relevant to the extent that everyone has a ROP. ROP is also criticized when applied to use of celebrity personae in expressive works. Is that a preference for amusement over discourse? There are only about 18 celebrity personality cases/year. That’s not nothing and litigated cases aren’t everything, but wants to focus on publicity interests of average citizens: the pleasurable servitude problem. Risk of identity loss means that “everyone belongs to everyone else,” as the slogan used in Huxley’s book goes. Class action ROP lawsuits against social media: result was broader consents in TOS. Voluntary relinquishment of identity control in return for the benefits of social media. Commodification of identity as a prerequisite for social media access. Social, political problem; social networks get monopolies over human capital.

Proposal: clickthrough policies designed to educate the public, maybe choices. 1A shouldn’t be a barrier to regulation b/c the use for endorsement is commercial speech.

Rothman: Does not agree that ROP is the coined opposite of the right of privacy, nor that it should have a purely economic and commercial focus. See her book. Also in her count there are 100s of ROP cases/year—order of magnitude more.

RT: Suggestion: Read Ashley Mears, Very Important People, on pleasurable exploitation and its relation to commodification and anti-commodification norms. Doesn’t have policy discussion itself, but has implications for solutions where individual relations seem pleasurable. Discussion seems indifferent to hidden data use; endorsement is almost literally the tip of the iceberg of individual data use. Proposal seems pretty weak tea; disclosure won’t work if they can still condition access on agreement.

A: On disclosure: Wants to be realistic about what could happen.

Wu: seems more unwitting [without thinking about it one way or another] and unavoidable transfer than pleasurable transfer. But in context of social media, the pleasure is inextricable from the agreement [and it’s not surprising that the agreement would then be experienced as, at least, not a problem].

Bita Amani, Authoring Identity: Copyright, Privacy, and Commodity Dissonance in the Digital Age

Emerging threats to capacity for self-authorship seem greater than in the past—here, algorithmic errors may generate disruptions in identity construction. Personal experience with multiple Bita Amanis with related interests. This has led to problems both with health care (corrected), misattribution of credit (interviews), and database connections as if they were all the same author.

Why should we care? Misappropriation; interference w/connection b/t author and text. Moral rights as a solution? Not clear. Peter Doig, well-known artist: denied creating a particular work; the owner claimed it was Doig’s work, and had to defend his identity to assert it wasn’t his work. The plaintiff pointed to style indicia of it being his. (Doig won; it was another guy named Peter Doige.) Privacy may have untapped potential for dealing with these misattributions, especially false light. Even true facts can be actionable; defamation is not required for intrusion on privacy/public disclosure of embarrassing facts. (Comes out of case involving nasty divorce where one party posted videos involving the kids on YT.)

Victoria Schwartz, Joint Privacy

[picking up kid; interesting project using ideas of joint authorship as a lens on issues of privacy that arise when people create information (or even just have information, as w/DNA) together and so sharing one’s own information or life story necessarily implicates others.]

Uri Y. Hacohen, User-Generated Data Network Effects

Network effects are key to current tech companies, whether via reviews, userbase, or otherwise. AI increases the power of network effects—making Google’s predictive results better. Many problems, including price discrimination, manipulation. Possible changes: changing liability regime so that they are more liable depending on what they know (e.g. that the user is a child); simple payment requirements to pay for harm [Pigouvian tax, I think]; management rights for users. Does not want to break up (at least as first solution) b/c that just means more entities with the data creating privacy and security problems, but we may not have a choice.

Felix Wu: Amazon, Google, and FB actually had different core businesses and if they’re all gulping this data then we have oligopoly, not monopolies. Even in a world of perfect competition, wouldn’t they be competing for who can manipulate best?

A: for FB, more data = more problems; right now they aren’t sharing as much as they might.

Wu: Some would say that innovation from the scale isn’t worth it; give up the marginal benefits and limit the size.

No comments: