Jylo has launched version 3.0 of its legal AI platform, adding configurable agents, automated workflow actions and controls over AI spending. The release gives legal teams more ways to turn AI-extracted information into work, while putting budgets and model access in administrators’ hands.
Watch Desk analysis
What happened
Jylo 3.0 is initially rolling out to early-adopter customer innovation and IT teams, according to Legal IT Insider. Its new Actions feature can use structured information extracted from documents to generate reports and presentations, redact content, assemble disclosure bundles or trigger bespoke workflows.
The release also adds a Library for building specialist agents, Outlook, PowerPoint and Excel add-ins, a secure code interpreter and expanded Model Context Protocol integrations. For document governance, Jylo says it has added version history and check-in/check-out, alongside deeper integration with iManage and NetDocuments. Administrators can allocate AI credits, set limits, control model access and manage usage by team or individual.
Legal IT Insider’s report on Jylo 3.0 includes comments from co-founder and CEO Shawn Curran and customers involved in the rollout.
Why it matters
The headline shift is from AI that analyses legal documents to software that can take the next step: producing a deliverable or triggering an action from extracted information. That could cut hand-offs in document-heavy work, while the new credit controls give firms a way to set boundaries on who can use which models and how much they can spend.
The detail matters more than the label “agentic”. These are features in an early-adopter rollout, not proof that every workflow is ready to run unattended. Version history and check-in/check-out also address the less glamorous, rather important question of how teams keep control of documents as AI gets involved.
Our read
This is a substantial product change, not merely a fresh coat of AI varnish. The combination of actions, specialist agents and spending controls gives legal teams concrete capabilities to assess. The real test will be whether firms can use them reliably in everyday work, with enough oversight to keep mistakes and costs from becoming someone else’s workflow.
What to watch
- Which Jylo 3.0 capabilities move beyond early adopters, and when.
- How firms govern agent actions and review the work they produce.
- Whether customers find the credit and model-access controls useful in practice.
- How the Excel and document-management integrations work in day-to-day legal tasks.
Discussion spark: Should legal AI agents be allowed to trigger actions such as redaction or disclosure-bundle assembly, or should every consequential step require a human sign-off?
Sources and evidence
- Exclusive: Jylo unveils ‘3.0’ with agentic workflows, Microsoft add-ins and AI spend controls – Legal IT Insider (13 October 2026, 13:01 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.