CodeX FutureLaw 2026: Dynamic Law for AI

Why It Matters

Dynamic law equips businesses and regulators to keep pace with AI’s rapid evolution, safeguarding consumer rights while unlocking new efficiencies and market opportunities.

Key Takeaways

• Law must become adaptable to rapid AI-driven societal changes.
• AI agents raise fiduciary duty concerns for consumer transactions.
• Escalation model—AI handles routine tasks, humans intervene for exceptions.
• Personal AI assistants can democratize access to government services.
• Open‑source tools enable transparent, loyal AI agents and flatter organizations.

Summary

The closing keynote by Stanford HAI fellow Sandy Pentland introduced the concept of "dynamic law" – a legal framework that can evolve as quickly as AI technologies reshape society. Pentland argued that traditional, slow‑moving statutes are ill‑suited for an environment where AI agents, open‑source models, and localized data processing are proliferating at breakneck speed. He highlighted three practical insights. First, AI‑driven agents create fiduciary‑duty dilemmas when they make purchasing or advisory decisions on behalf of users. Second, the most successful AI deployments follow an "escalation" pattern: routine tasks are automated, while humans intervene for novel or complex cases. Third, personal AI assistants can dramatically improve access to government services, as demonstrated in a Washington‑DC town‑hall where single‑parent respondents asked for AI help filling forms and tracking benefits. Pentland illustrated his points with concrete examples: a startup he co‑founded that automated immigration paperwork (later acquired by Cleo), the LoyalAgents.org project with Consumer Reports to certify agent loyalty, AI‑facilitated meeting summarizers that doubled satisfaction scores, and a Davos round‑table envisioning virtual legal advisors that crowdsource expert opinions at a fraction of traditional costs. The implications are profound. Dynamic, AI‑compatible law could flatten corporate hierarchies, enable decentralized collaboration, and spawn new business models around transparent, loyal agents. At the same time, regulators and firms must grapple with fiduciary responsibilities, interoperability across jurisdictions, and the ethical limits of delegating decision‑making to machines.

Original Description

The law is permanent by design — but the world is changing faster than ever. How do we fix that? In this closing keynote, MIT/Stanford professor Alex “Sandy” Pentland argues that AI doesn't just change what lawyers do — it changes the fundamental architecture of how law is made, interpreted, and applied. From AI agents with hidden loyalties buying things on your behalf, to virtual legal experts you can consult for $500 a ping, to flattened law firm hierarchies where small teams serve millions, the transformation is already underway. Pentland also reveals something surprising: most AI deployments in large organizations don't reduce headcount — they expand what's possible. And for the millions of people who currently can't access justice or navigate government bureaucracy, a personal AI that fills out your forms, tracks your deadlines, and tells you when the law changes might be the most important legal innovation of our lifetime. This is what dynamic law looks like — and it's more human than you might think.

Alex “Sandy” Pentland is HAI Center Fellow and faculty lead for digital society at Stanford HAI and Digital Economy Lab, He is Toshiba Professor at MIT, member of US National Academies, Advisor to Abu Dhabi Investment Authority Lab, and formerly advisory board member at UN Secretary General’s office, Google, ATT, Telefonica, and elsewhere.