Legal tech M&A is no longer about tools, but about talent and training
Legal tech M&A is the trend of established legal education and information companies buying AI startups to control how artificial intelligence is embedded into lawyers’ training, tools, and everyday workflows, rather than leaving that transformation to third‑party vendors or ad‑hoc experiments.
BARBRI’s acquisition of legal AI startup Lega is a clear example: a traditional legal education business buying an AI platform specifically to offer experiential AI workshops, simulations, hackathons, and lab‑style learning experiences at scale. At the same time, Thomson Reuters has opened early access to the next generation of its CoCounsel legal AI assistant and is testing its own large language model for legal tasks, called Thomson. These are not side projects; they are signals that the center of gravity in legal education AI and legal practice automation is shifting inside the platforms that already manage professional training and research. If you are a lawyer or law student, your future AI habits will be shaped less by open web tools and more by whatever stack your institution has bought.
BARBRI + Lega: experiential AI training as the new bar prep
BARBRI did not buy Lega to pad its course catalog; it bought a sandbox. The legal education provider has acquired Lega as part of a strategy to deliver experiential AI workshops, simulations, hackathons, and lab‑style learning experiences. Lega’s SaaS platform was built to give legal organizations a safe, structured environment to experiment with generative AI, compare models, build AI‑powered tools, and explore practical use cases. Now, that same environment is being repurposed for systematic training rather than piecemeal experimentation.
Through the deal, Lega’s experiential learning model can now be delivered at scale to law students, lawyers, law firms, and legal organizations worldwide. BARBRI’s co‑CEO describes the platform as a sandbox where learners can experiment, build things, build confidence, and build AI fluency, slotting it alongside existing AI courses and CLE offerings. The plan is explicit: take these sandbox environments into law schools, run workshops and hackathons, and bridge the gap between law school and law firm by letting people get their hands dirty with practical application. In this model, the AI legal assistant is not a gadget students stumble onto; it is the classroom itself.
Thomson Reuters and CoCounsel: from plugging in AI to owning the stack
If BARBRI’s move is about training, Thomson Reuters’ strategy is about infrastructure power. The company has opened early access to the next generation of CoCounsel Legal, calling it the most substantial reworking of its flagship legal AI assistant since acquiring Casetext in 2023. This new CoCounsel is the first time the company says it has truly unlocked the value of its content and the expertise of more than 2,700 attorney‑editors in training the agent. In other words, the AI legal assistant is now tightly coupled to proprietary legal content and editorial workflows, not sitting on top of them.
The company has built its agentic AI products to be model‑agnostic and has partnered with Anthropic’s Claude models, but it has also developed its own large language model for legal, imaginatively called Thomson. That model is starting to outperform the latest alternatives on specific legal tasks, and the company may run some or all aspects of CoCounsel on Thomson while keeping other parts on Claude. This is legal tech M&A evolving into AI infrastructure sovereignty: owning Casetext, owning the content, owning the LLM. The message to the market is blunt: if you want fiduciary‑grade AI that is the opposite of a black box and walks users through every step with citations and reference material, you will find it baked into the research platforms you already pay for.
Consolidation, control, and the new AI‑ready lawyer
Look at the pattern: BARBRI buys Lega, Thomson Reuters acquired Casetext in 2023, and both immediately weave AI into their core products instead of leaving it as an add‑on. This legal tech M&A wave is about control of innovation. Established players are not content to integrate third‑party AI; they are buying the startups, the platforms, and in some cases the research labs so they can dictate the pace and shape of legal practice automation. According to Thomson Reuters’ Future of Professionals Report 2026, there is already a widening gap between AI adoption and the value organizations extract from it. Owning the end‑to‑end AI stack is their answer to that gap.
At the workflow level, AI‑powered legal tools now handle complex research, contract analysis, and other specific legal tasks across structured sandboxes and agentic assistants. Lega’s platform helps law firms safely explore, assess, and implement generative AI technology in practice, while CoCounsel’s new version is built to exploit proprietary content and editorial expertise for higher‑accuracy outputs. That shifts what “AI‑ready” means for lawyers: the key skill is no longer knowing how to search; it is knowing how to design prompts, evaluate AI‑generated reasoning with citations, and decide when to trust or override automated workflows. The next generation of legal professionals will be trained inside these controlled AI environments—and that is by design.
Conclusion: AI training is becoming the new curriculum, not a side course
The most important change in legal tech right now is not a single product launch; it is who owns AI literacy. BARBRI is turning AI sandboxes and hackathons into a structured part of legal education, putting experiential AI learning on the same footing as exam prep and CLE. Thomson Reuters is binding its AI legal assistant, CoCounsel legal AI, to its own content and large language model, positioning AI as the default interface to its research ecosystem.
This consolidation should make legal education AI more coherent, but it also concentrates power. When a handful of platforms control both the training and the tools, they define what “good” AI practice looks like and which skills matter. For practitioners, the response cannot be passive. If AI‑driven workshops, simulations, and assistants are becoming the core curriculum, lawyers must treat them as a craft, not a shortcut—insisting on transparency, testing limits, and building judgment on top of automation rather than handing it over.
