iManage at ILTACON: why governed context beats the model
James Ainsworth, global head of content at iManage, outlines why data, context, and governance are key factors that determine how much value AI brings and shares insight from ILTACON
AI is already changing the shape of legal and knowledge work. The organisations that are able to realise the value of AI at scale must focus on more than just the model.
In his keynote at ILTACON 2026, Neil Araujo, CEO at iManage, pointed to Gartner research that puts a number on why the data behind that AI matters as much as the AI itself.
As Rita Sallam, distinguished VP analyst at Gartner, has previously shared: “Organisations that fail to adopt comprehensive context structures — supported by a robust data layer — will perpetuate data inefficiencies and face heightened financial costs, as well as legal and reputational damage.”
Get the data right, and Gartner’s research points to gains of up to 80% in model accuracy and a 60% reduction in costs. Freed-up hours only become client value if the knowledge behind them compounds. For law firms, a matter handled well this year should make the next one faster, not force every team to start from zero.
Two things determine whether AI delivers that value: context and governance
For a period, the model itself was the industry’s proxy for AI quality, but that phase has largely passed. Araujo named where the real differentiation lives now: “It’s not the model. It’s not the layer that goes above the model. It’s your data, your knowledge, your experience — that’s what makes you different.” Every organisation now has access to broadly the same frontier models. What separates one firm’s AI assistant from another’s is what it’s allowed to see, and whether its answers can be traced back to something trustworthy.
That’s a governance question as much as a technical one. An AI system validated against an organisation’s own approved guidance, with every answer traceable to its source, lets people apply judgment with confidence rather than checking AI output against nothing at all.
Governed context has to work at scale, not just in theory
Araujo walked through what the iManage context fabric™ draws on: the documents and communications at the centre, activity signals showing who worked on what and which version matters, and metadata that identifies what a document is and how current it is. None of that is enough on its own. He gave a concrete example: negotiating the same clause with the same opposing counsel for the tenth time and needing to know their position on a point like limitation of liability from the last negotiation. That information doesn’t only live inside the document. It also lives in a practice management system (PMS), and only a connected, governed layer can surface it when it matters.
In his own ILTACON keynote, Shawn Misquitta, EVP at iManage, extended the point with a second example from the platform’s expert-finder capability: rather than searching lawyer bios for someone who has drafted a facilities agreement before, the system cross-references who has actually worked on comparable deals and narrows to the right two or three people. Context, in other words, extends beyond documents to firmwide expertise.
The next-gen platform: where governed context turns into client value, at scale
This is where the next-generation iManage platform comes in. iManage product strategy is aligned to three themes: a governed core, context that explains what’s relevant and why, and connected delivery wherever the work happens. All three arrive together when the platform reaches general availability this October — read the full announcement.
The platform now extends governance to external collaborators and contractors without breaking the single source of truth or its audit trail, and adds disposition tools to retire redundant, obsolete, or trivial content at scale. This is because AI reasoning over messy files and clutter is more likely to produce inaccurate or irrelevant answers.
Governance now also extends explicitly to AI itself: policies can specify whether a given client’s data may be used with AI at all, and with which tools. That addresses outside counsel guidelines and client restrictions raised directly by customers, including government engagements and patent work, where firms are wary of what a model retains.
The long view
2026 marks the 30th ILTA appearance for iManage — a company now in its 31st year. Araujo closed on this: a platform built by an organisation that thinks in decades is now shipping the foundation for scaling AI into client value. Not promising, shipping. General availability follows in October.
