Algorithmic Harm: A Response to Critics

Abstract In our 2025 book, Algorithmic Harm, we contend that the use of algorithms can benefit consumers in many ways (largely because of greater personalization), but that it can also cause harm in the face of information deficits and behavioral biases (again, largely because of greater personalization). Unsophisticated consumers, as we call them, are especially vulnerable to algorithmic harm. In this short response to a set of excellent comments on our book, we explore some of the benefits of personalization; some of the costs of forbidding it; some challenges, in terms of feasibility, to our preferred approaches; and the intriguing question whether and in what sense algorithms might be said to have an unconscious.