You know the request. A partner wants a new AI tool by Friday. A client wants to know where its documents go. The CFO wants to know why the AI line keeps growing. And somewhere in the firm, an associate is pasting a draft into a tool you have never heard of.
Say yes, and you own the risk. Say no, and you become the reason the firm is falling behind. Either way, you are the one who answers when something goes wrong.
That is not a technology problem. It is an economics problem that lands on your desk because nobody else can see the whole picture. Consider five facts.
1. Shadow AI is already inside the firm.
Thomson Reuters reports that roughly a third of surveyed law-firm professionals use AI tools their firms have not authorized. The tools you approved are not the only ones processing client work. They are just the only ones you can see.
2. Saying no is no longer a strategy.
In Clio’s 2026 research, 85 percent of legal professionals reported some AI adoption at their firms. Blocking a tool does not stop the demand for it. It moves the demand somewhere you cannot govern. The firms that stay safe will be the ones that make the approved path the easiest path.
3. Your clients now write the rules.
Security, ethics, privilege and trust consistently rank among the top barriers to AI adoption in legal-industry surveys. Clients are turning those concerns into conditions: which matters may use AI, which providers are acceptable, and how their data is handled. Compliance is no longer a firm-wide policy. It is a matter-by-matter obligation.
18%
of professionals across professional services say their organization measures the return on its AI investment.Thomson Reuters, 2026
4. AI costs don’t compare cleanly.
Seats, tokens, gateways, bundled features and privately hosted models are billed in different units, by different vendors, into different budgets. Putting two options side by side is hard. Explaining what either one did for a specific client is harder.
5. When someone asks for the return, they will ask you.
So few organizations measure AI’s return today that the question is still mostly unasked. It won’t stay that way. When the managing partner wants to know what the firm got for its AI spend, the CIO who can answer with evidence earns a seat at the strategy table. The one who can’t gets a budget review.
What we have built, and why we will not describe it here.
mattermeter was founded to answer a question every law-firm CIO will soon be asked: what is AI really costing the firm, and what is it worth? We have built something specific. We are not describing it on a public website. You of all people will understand why.
What we will say is this. It was built for finance, with your IT. It treats privilege and client confidentiality as non-negotiable. It works alongside the systems you already run rather than asking you to replace them. And it expects your security team to review it as thoroughly as any vendor should be reviewed.
Here’s what I’d like you to do.
Fill in the short form below. It takes less than a minute.
I will reply personally and arrange a private technical briefing. Bring your CISO or your head of security if you like. We will talk about your environment, your approval process and the questions your partners and clients are already asking. You leave with a clearer view of your AI economics, whether or not we ever work together.
Now the part I have to be straight with you about.
We are taking on no more than [5] founding firms before launch. That is how many we can serve properly while we are this small. Founding firms see the work first, help decide what is built, including how it fits your security model, and get terms we will not offer again. When those places are gone, they are gone.
The shadow tools aren’t waiting for next quarter’s budget cycle. Neither should you.
Sincerely,
[Founder name]
Founder, mattermeter
P.S. If your firm has a standard security questionnaire, mention it in your request. We would rather answer it before the briefing than after.
P.P.S. We wrote a different letter for your CFO. It asks who keeps the hours AI saves. You may want to read it together.