Algorithmic surveillance: the collection conundrum

ABSTRACT

Supporters of increased surveillance see tremendous potential in the ever increasing creation, collection, and retention of personal data. Most acknowledge that the massive collection of information also creates challenges where the collection outpaces the ability to meaningfully process the data. Increased processing power and more finely tuned algorithms are often portrayed as the solution to this haystack conundrum. While a human may struggle to find the needle in an overflowing haystack of disordered information, powerful computers can take a logical and structured approach that will make the haystack eminently more searchable. This article evaluates this premise from a human rights perspective and considers whether algorithmic surveillance systems can be designed to be compatible with the right to privacy. In addition to assessing the incongruity between traditional safeguards (such as foreseeability and accountability) with algorithmic surveillance, this article also confronts the problem of initial collection and addresses the contention that well-defined algorithmic search can effectively limit the intrusiveness of surveillance. Evolution in the case law of the European Court of Human Rights and the Court of Justice of the European Union will be factored into this analysis.