Buyers at law firms, like chief innovation officers or CIOs, are getting buried in pilots to make sense of the current state of the art and equally buried in the demands for AI that clients and partners are bringing with urgency. AI is still progressing rapidly, with new breakthroughs and advances on a monthly basis, and the way that inflects how legal tech is racing to keep up creates new crests and valleys across categories of legal AI. Most of all, we observe a convergence of building, buying and using AI in law, a topic we’ll discuss more in the market in coming months.

But the market map is far from mature, and with the convergence comes category blur – buyers are no longer talking about document management, knowledge management, transactional workflow automation, drafting tools for litigation, and legal research as discrete categories. The questions feel more like apples to cheeseburgers: do we choose Harvey, or do we build our own proprietary AI system? Or do we lean into a partnership with OpenAI or Anthropic, or do we utilise the AI being built into our DMS?
And it’s important to remember that it wasn’t always like that. While some solutions always pushed the bounds of categories, a law firm or legal team could put together an RFI and identify a set of leading players that they should consider. A buyer could send someone to a trade show (like LegalTechTalk), have them focus on a single wing of the exhibit hall, collect collateral and see demos of a set of tools, then come back with an impression. Today, that’s a nearly impossible task.

Unlike previous cycles of technology disruption, firms must act – the cost of inaction is the greatest risk. So, the answer will sound obvious: buyers must pay attention to their own users, their clients, their assets against data and talent, and the strategic priorities of their firm. Once that is well understood, they can narrow the scope and shorten the cycle to identifying, piloting and buying.