By Caroline Hill
One of the most interesting themes of 2026 so far for me has been an increased focus on the need for organisation-wide, governed orchestration, and a growing awareness that individual productivity tools will not of themselves lead to systemic change.
Much of the first wave of generative AI in legal has focused on productivity: making a single lawyer faster or reducing friction in discrete tasks. What it has mostly not done, at least not yet, is meaningfully rewire the operating model of legal work. In practice, firms have bolted AI tools onto existing processes, often creating new silos rather than removing old ones. The result is often speed in pockets, organisational inefficiency, and lack of auditability and control.
The company that prompted me to start thinking about governed orchestration recently is Autologyx, and more specifically, a conversation with chief commercial officer Ben Wightwick. Autologyx focuses on legal and compliance workflows and their argument is that the next phase of innovation is not just about smart tools, but where those tools live; how they are leveraged in a controlled way; and how they are audited when AI begins to participate in work product delivery itself.
Autologyx provides an integration framework to connect systems, tools and workflows. In a world where everything moves at lightning speed, the argument is that organisations need at their core a structured layer into which AI can be slotted, governed and controlled. In other words, innovation needs a spine.
This matters because legal work, particularly in regulated and risk‑sensitive contexts, cannot be probabilistic end‑to‑end. Firms need to know what happened, when, and why. They need determinism, auditability and hand‑offs between systems, people and decisions. AI can add intelligence at specific points in a process – classification, extraction, drafting, triage – but it cannot, on its own, provide the accountability that legal work demands. As Autologyx asks in a recent blog post, how do we scale AI without scaling risk?
A common example is large‑scale contract remediation, Wightwick says. Firms may already use AI tools to review hundreds of agreements, but stitching together review outputs, risk decisions, client logic, document generation and regulatory reporting often still involves multiple systems and a great deal of manual effort. The inefficiency is not in the AI; it is in the lack of an underlying process architecture.
What’s emerging across the market is not resistance to AI, it’s a governance gap. AI tools optimise for flexibility. They empower individuals and encourage exploration. Legal delivery, by contrast, demands consistency.
What Autologyx is effectively advocating is a shift in focus towards how to embed AI into standardised ways of working, requiring integration with document management systems, clear business logic, human‑in‑the‑loop controls, and the ability to swap AI models in and out as they evolve.
Adoption problems are often blamed on lawyers, training or culture. But if AI is delivered as a bolt‑on productivity aid, resistance is almost inevitable. When it is embedded into the workflow itself, change becomes less about persuasion and more about design.
The legal sector is moving past the press‑release phase of AI. The more difficult work now is building systems that can absorb innovation without destabilising the practice of law. If productivity was the headline of the first wave, infrastructure may be the quieter, but more consequential, story of the next.
See also:
From TAR TO HAR Report: Are GenAI Discovery Tools Ready for Forensic Scrutiny? Read it here now!
The post Giving innovation a spine: Why organisations need governed orchestration appeared first on Legal IT Insider.