AI Compresses Production Cost Unevenly
Not every patent task is affected in the same way. Formatting, translation assistance, document comparison, prior-art triage, prosecution-history summarization, and first-pass claim mapping can often be accelerated substantially. Claim construction, materiality, inventorship, amendment strategy, FTO risk acceptance, and final legal conclusions are harder to compress because they depend on context and accountability.
This creates a margin between the cost of producing information and the value of making the right decision from that information.
Hourly Billing Has a Built-In Efficiency Tension
For lawyers who charge hourly, professional-responsibility rules still matter. ABA Formal Opinion 512 states that hourly billing must reflect actual time spent. AI cannot be used as a hidden mechanism for billing the hours a task used to take.
The commercial consequence is straightforward: if a firm becomes more efficient but sells only time, its revenue per task can fall even as its capability improves.
What Can Replace Pure Time Pricing?
There is no single answer. Depending on jurisdiction, engagement type, professional rules, and client expectations, firms and expert practices may use structures such as:
- fixed fees for clearly defined deliverables
- phase-based pricing tied to decision gates
- subscriptions or retainers for recurring portfolio support
- tiered review levels based on risk and required seniority
- project pricing for benchmark design, AI evaluation, or expert QA
- hybrid structures combining a fixed scope with hourly work for defined exceptions
The objective is not to conceal efficiency. It is to make scope, quality level, responsibility, and price legible to both sides.
What Clients Will Still Pay a Premium For
AI reduces the scarcity of first drafts. It does not reduce the scarcity of reliable judgment at the same rate. In patent work, premium value increasingly sits in activities such as identifying the decisive claim limitation, rejecting a false prior-art mapping, recognizing a prosecution-history trap, designing a viable fallback position, or telling a product team that a technically elegant workaround does not actually avoid the patent.
These are exactly the places where an experienced reviewer can prevent a cheap output from becoming an expensive decision.
The Same Economics Apply to AI Subject-Matter Expert Work
AI companies also need domain experts not merely to produce more text, but to create evaluation infrastructure: realistic tasks, gold answers, rubrics, error taxonomies, adjudication protocols, and high-confidence review. Those deliverables are valuable because they improve model reliability across many future outputs.
That creates a different unit of value from traditional patent production. The client is buying a reusable standard of correctness, not a stack of billable pages.
Takeaway
AI does not eliminate professional pricing. It exposes weak pricing logic. Practices that rely on routine production hours will face pressure. Practices that can define, verify, and take responsibility for high-value decisions have more room to move toward transparent, outcome-oriented scopes and prices.