# a16z’s David Haber Outlines Revenue Models for Long-Running AI Agents

By Simon Yoon

Canonical URL: https://www.tokenpost.com/news/technology/32141
Published: 2026-10-15T00:38:25.000Z
Updated: 2026-10-15T00:38:25.000Z
Section: Technology

> The agents could handle work across weeks or months, including steps in insurance, mortgages and plaintiff law firm case preparation.

AI agents that manage work over weeks or months could support business models tied to revenue growth, a16z General Partner David Haber said, describing possible uses in insurance, mortgages and plaintiff law firm case preparation.

Haber said these agents can handle dozens of steps and different forms of interaction. AI had previously handled mostly short tasks with clearly defined boundaries, while customer service advanced sooner because results are easier to verify and people can take over unusual cases.

Long-running agents could guide an insurance customer from an initial call through payment, move a mortgage from preapproval to closing, or help advance a business lead from a product demonstration to a signed contract. The model could reward revenue growth rather than focus only on cost savings.

Haber pointed to plaintiff law firms as an example. Prospective clients may contact several firms, and a slow response or evaluation can cost a firm an opportunity. After signing, limited staff capacity can constrain evidence gathering, including follow-up with witnesses and review of lower-value cases. Missing bills or unexplained gaps in treatment can reduce settlement amounts.

Agents could assess inquiries at any hour and use subagents to advance evidence gathering and client updates by phone, email or text. They could also maintain records ranging from call logs to medical documents.

Haber outlined three commercialization paths: selling a platform to existing teams, taking on services those teams already outsource, or combining the technology with the service. As an example of the third approach, he cited buying debt portfolios and handling collections.

Most U.S. states prohibit nonlawyers from sharing legal fees. In plaintiff-side legal work, pricing can be aligned with settlement amounts, but cannot be structured as a direct share of legal fees.