Problem: Your Law Firms Use AI. Your Invoices Didn't Notice.

The problem

The numbers are consistent across surveys:

• In the ACC/Everlaw GenAI survey, 59% of in-house respondents said they had seen no noticeable savings from their law firms' use of GenAI, and only 24% were satisfied with how outside counsel used it to improve cost-effectiveness.
• 61% plan to push for pricing changes tied to AI.
• At Legalweek 2026, Thomson Reuters observed that client expectations have shifted "from curiosity to accountability": clients now want proof of savings on specific matters, which most firms still can't provide.
• Meanwhile, law firm rates grew more than 7% in 2025 and are expected to keep climbing.

In September, the conversation reached a boil, with reports describing firms and clients actively battling over who gets to keep the AI dividend.

Under the billable hour, AI efficiency doesn't lower your bill. It lowers the firm's hours, and then the firm raises the rate.

The diagnosis

There are usually three reasons savings don't appear.

1. The pricing model eats the savings

If a task that took six hours now takes two, hourly billing should cut the fee by two thirds. In practice, that time gets absorbed: more review, more polish, higher rates on the remaining hours. The efficiency is real; your share of it is not.

2. You can't see the work

Line-item invoices describe activities, not outcomes. Without consistent task coding and matter data, you can't compare "contract review, 2025" to "contract review, 2026" and prove anything.

3. Nobody asked in writing

If your outside counsel guidelines don't address AI (disclosure, billing for AI-assisted time, who bears tool costs), then the default rule is whatever the firm prefers.

The solution

Get the data first: e-billing and spend analytics

You can't negotiate what you can't measure. An e-billing and spend management platform standardizes task codes and enforces your billing guidelines automatically. Layer on spend analytics to benchmark cost per matter type, staffing mix, and cycle time year over year. When you can show that the average NDA review still costs the same as in 2024, the conversation changes.

Put AI in your outside counsel guidelines

Add a short AI section to your OCGs:

1. Disclose material use of AI tools on our matters.
2. Don't bill for time spent learning or configuring AI tools.
3. Don't pass through AI tool costs without prior approval.
4. Reflect AI efficiencies in pricing for repeatable work.
5. Protect our data: no training on our information, and approved tools only.

Move repeatable work off the hourly model

For high-volume, predictable work (NDAs, standard commercial contracts, routine employment matters), ask for fixed fees or portfolio pricing. Under a flat fee, AI efficiency becomes the firm's margin and your predictability, which is a deal both sides can live with.

Ask better questions at the next review

• "Which of our matters used AI this year, and what did it change?"
• "Show us cycle time and staffing for our top three matter types versus last year."
• "What would a fixed fee look like for this category of work?"

Consider bringing some work home

Sixty-four percent of in-house teams in the ACC/Everlaw survey said they expect to rely less on outside counsel because of AI. If your own AI assistant and contract tools can handle the first pass, you may not need to send it out at all.

The bottom line

AI savings from outside counsel won't arrive on their own. They show up when you have the data to measure them, the guidelines to require them, and the pricing model to capture them. Start with your billing data. Browse e-billing and spend management tools on CorporateLegal.tech to see who can get you there.