Lawfare Archive: Robert Mahari on Addictive Intelligence, Digital Attachment Disorders, and Other AI-related Concerns

The Lawfare Podcast features Robert Mahari, a joint JD-PhD candidate at MIT Media Lab and Harvard Law School, discussing the psychological and social risks posed by increasingly sophisticated AI agents. Hosted by Kevin Frazier (Assistant Professor at St. Thomas University College of Law and Tarbell Fellow at Lawfare), the 37-minute episode explores how prolonged interaction with AI systems may foster addictive or unhealthy attachment patterns in users. Mahari draws on his recent MIT Technology Review piece (co-authored with Pat Pataranutaporn) to examine current uses of generative AI and the potential for what he frames as 'digital attachment disorders.' The conversation appears to center on the mechanisms by which AI agents—designed to be responsive, personalized, and engaging—could exploit human psychological vulnerabilities, particularly around social connection and dependency. The episode positions this as a legal and policy concern, fitting Lawfare's focus on governance and institutional implications of emerging technology. While the show notes do not detail specific examples, case studies, or policy recommendations discussed, they indicate the episode addresses both the theoretical framework (addiction and attachment psychology applied to AI) and practical concerns (how existing generative AI systems are already being used in potentially problematic ways). The episode is framed as timely commentary on a nascent but urgent issue as AI agents become more prevalent.

Verdict: A timely, credibly-sourced interview on an emerging concern, but limited show notes prevent confirmation of depth, balance, or actionable takeaways.

Government rubric

• Accuracy and sourcing 3/5 20%

  Guest has credible credentials (MIT Media Lab/Harvard Law JD-PhD) and published in MIT Tech Review, suggesting sourced research, but show notes don't detail specific claims, evidence, or citations discussed.

• Balance and context 2/5 20%

  Show notes frame the topic as a concern/risk without indicating whether counterarguments, industry perspectives, or alternative interpretations are presented.

• Clarity on what it means 2/5 15%

  The real-world policy or practical implications of 'digital attachment disorders' are not explained in the show notes; unclear what listeners should do with this information.

• Substance 3/5 15%

  Guest expertise and published research suggest substantive content, but 37 minutes is relatively brief for a complex topic; show notes don't confirm depth or reveal whether discussion stays focused.

• Delivers on its title 4/5 15%

  Title promises discussion of addictive AI, digital attachment disorders, and AI concerns; show notes confirm all three topics are addressed by a qualified expert.

• Value for time 3/5 15%

  37 minutes is reasonable for an interview format, but without knowing pacing or tangents, moderate rating is appropriate; length seems adequate for the topic scope.

How we score Government

Worth it if you're interested in AI governance, psychology of human-AI interaction, or emerging policy issues around AI agents, and don't mind a relatively introductory treatment.

Based on the episode description. AI-generated; it can be wrong.