Will Chen, a developer, saw enterprise legal AI demos and realized their premium pricing wasn’t justified by basic features like chat interfaces or prompt templates. By building and releasing MikeOSS, he showed that much of what is marketed as sophisticated legal AI can be reproduced by skilled individuals.
When Will appeared on a fireside chat with Jamie Tso and Raymond Sun of Legal Quants shortly after, the conversation moved quickly to what MikeOSS actually revealed. Three bargaining power shifts are happening at once. Each follows from the same underlying move: someone who was assumed not to understand what they were buying figured it out. That happened first between firms and vendors. It is now happening between innovative lawyers and the firms that employ them, as well as between clients and outside counsel. The client shift is the one most firms are not watching closely enough. When an in-house team runs the same calculation Will ran, watching a demo and asking what it would cost to build, the case for sending routine work outside gets harder to make.
What You Are Actually Paying For
Vendors charge what they charge because buyers have not been able to evaluate what they are buying. MikeOSS changes that. The chat interfaces, the playbooks, the document review tables: those are replicable, as Will demonstrated. What companies like Harvey and Legora legitimately earn their fees for is the enterprise wrapper: security, deployment, configuration, and support. That is a service business, not a moat. Before your next renewal, the question worth asking is which of those two things you are actually paying for, and whether the price reflects it.
The next round of legal AI value will not come from general platforms. Will was clear about this in the fireside chat: generalist tools will be copied, and the firms that build on top of them for a specific workflow or jurisdiction will be the ones creating defensible value. That applies to vendors building on MikeOSS, and it applies to the lawyers inside your firm who understand a workflow well enough to improve it. General capability is becoming a baseline. Depth is where the advantage will be.
The Harder Problem Is Inside the Firm
The harder challenge is inside the firm, and this is where you need to be honest with your partners. Most compensation systems reward production. They do not reward the creation of tools that make other lawyers more productive. If you want that to change, build incentives around what you actually want to reward. A lawyer who makes twenty other lawyers more productive has created real value for the firm. Integrate that recognition into formal evaluations alongside billable production.
Consider this: If a junior lawyer created a tool that saved forty hours on a fixed-fee matter, how would your firm reward them? Typically, origination credit goes elsewhere, and fewer billable hours may even penalize the innovator. Law firms lack an equity-sharing system like those used by software companies.
Will described how most firms reward output rather than the creation of tools that boost others’ productivity. Jamie pressed him on this misalignment. The implication of that conversation is direct: if your firm wants real innovation, the compensation system has to recognize it.
What Happens to Your Billing Model
Jamie raised pricing directly: If AI increases lawyer productivity, what happens to billing? Fewer hours mean the rate-times-hours formula works against you. Move to value-based or fixed pricing before clients require it. Most firms haven’t. They add AI subscriptions but keep billing unchanged. Clients will soon get better tools, making this model hard to justify.
Raymond Sun pressed Will on this directly. Will’s answer was that client relationships still matter, but clients will increasingly want measurable results from AI. Right now, saying the firm uses Harvey is no longer a differentiator. Clients will want to see concrete outcomes. A firm that cannot show what its AI investment produces is competing on a claim that the whole market is already making.
Raymond Sun’s questions pushed toward the strategic position of law firms and in-house teams. If AI-native firms charge premium prices, where does the money come from? If client legal budgets remain constrained, will in-house teams use open-source tools to do more themselves?
Will agreed that this is a real possibility. In-house teams may not replace outside counsel for complex transactions or high-risk litigation. But they may use open-source tools and enterprise AI subscriptions to handle more repeatable work internally.
The Real Lesson
This is why MikeOSS matters. It is not only a product. It is a signal that the cost of building is falling and that law firms should no longer treat legal AI as a black box.
Commercial legal AI platforms will still matter. Many firms will prefer supported, secure, enterprise-ready tools. But the existence of open-source alternatives should make the market more honest. Vendors will need to show where their value really sits. Firms will need to understand which workflows are worth buying, which are worth building, and which should be redesigned altogether.
The firms that benefit most from AI will not necessarily be the firms that buy the most impressive platform. They will be the firms that understand their own work deeply enough to know where AI can create economic value.
That is the real lesson of MikeOSS.
Legal AI strategy cannot simply be a software purchase. It touches pricing, compensation, governance, client service, training, and profitability. The firms that understand that will have more bargaining power than the firms that simply buy what they are sold.
