Skip to content

Ethics & Neurorights

When LLMs can reconstruct neural activity into text and images, "mind reading" is no longer science fiction. This raises a fundamental legislative question: are signals in the brain more private personal data than DNA?

Since 2021, Chile, Colorado, Minnesota, UNESCO, and the EU have successively proposed legislative frameworks for "Neurorights." This chapter discusses the philosophical stances behind this legislation, specific provisions, and the unique risks brought by the LLM + BCI combination from an AI alignment perspective.

Why this chapter sits here. Chapter 07 (brain-to-language), Chapter 08 (brain-to-image), and Chapter 12 (consumer EEG) have already turned "mind reading" from theory into a capability you can buy off the shelf. Once capability is in place, the legislative window closes fast — so this chapter is not bonus material but a constraint that engineers should weigh on equal footing with feasibility when making technical choices. It directly addresses the compliance boundaries of Chapter 12 (consumer EEG data collection), Chapter 11 (clinical BCI data protection), and Chapter 08 (visual reconstruction can recover private visual memories).

Recommended reading order. Start with Neurorights & Cognitive Liberty Legislation to master the specific provisions in five jurisdictions (Chile 2021 constitution, Colorado 2024, Minnesota 2024, UNESCO 2024, EU AI Act) and how each defines "neural data." Then Brain Data Privacy & Cognitive Biometrics takes in Ienca et al.'s argument for why brain data cannot simply be retrofitted into the existing GDPR / HIPAA frame. Next, AI Alignment Perspective examines the unique risks of LLM-based mind reading — misalignment of objectives via neural supervision and the design of user-control mechanisms. Finally, Governance Framework: UNESCO / EU AI Act provides the compliance pathway for actually shipping a product.

Chapter contents:

评论 #