In this article we explore elements of the Sourcing & Delivery pillar of the Target Operating Model as introduced in our series launch: "AI for in-house legal teams: five pillars for success".

We focus on how AI enables legal teams to fundamentally rethink their delivery models, moving beyond incremental improvements to create systematic frameworks that match work types to optimal delivery channels.

From incremental to transformational

Most in-house legal teams have evolved their delivery models over time; adding a self-service portal here, some templates there, perhaps a panel of preferred external firms or even some Alternative Legal Service Providers. These incremental improvements have helped, but they haven't fundamentally changed how legal work gets done or where value is created.

AI has created entirely new delivery options, not just improved legal tools. Work that previously took a significant amount of lawyer time can now be handled by the business with AI guidance and legal oversight. Work that once took days of lawyer time can be pre-screened, analysed and summarised by AI in minutes. Most importantly, matters that fell into a grey zone - too complex for pure self-service but too routine to justify full lawyer attention - now have a natural home in AI-augmented delivery, eliminating the false choice between 'lawyer does everything' and 'business does it alone'.

The result? Rightsourcing - where legal teams can now operate across a true spectrum of delivery models, matching each type of work to the resource that delivers the best combination of speed, quality, cost and risk management.

For example, imagine a legal team that redesigns its delivery model using a tiered framework leveraging AI across multiple channels. Standard contracts that once took a week for lawyer review are now handled through AI-enabled self-service in hours. Commercial agreements are pre-screened by AI, with lawyers focusing only on genuine issues and negotiation points. Lawyer satisfaction increases as routine work disappears, and strategic matters get proper attention. The transformation isn't just about efficiency. It fundamentally repositions legal from bottleneck to business enabler.

The opportunity is clear. The question is: do you have a framework for seizing it, the capabilities to deliver it and the capacity to execute it?

A framework for optimum delivery

The best legal teams don't route work based on tradition or instinct. They use systematic criteria to match each matter type to the delivery model that optimises speed, quality, cost and risk management. Without this discipline, you'll default to old patterns; routing everything through lawyers regardless of complexity, or pushing work to self-service without proper guardrails.

The framework starts with three fundamental questions, noted in our previous article AI for in-house legal teams: start with legal strategy, which should guide the decisions you make:

• What work should the business deliver themselves, supported by AI, approved and overseen by the in-house legal function?
• What work should a lawyer deliver with a combination of human judgement and AI?
• What work needs to remain the domain of pure human expertise?

The Four-Criteria Assessment

Answering those three questions in principle is straightforward. The challenge emerges when evaluating your actual portfolio, which includes dozens of different matter types, each with vocal stakeholders and established workflows. How do you ensure consistency across evaluations? How do you defend your decisions with evidence rather than intuition?

The solution lies in systematically assessing key criteria to determine how work should be allocated or prioritised based on your specific operating context. By evaluating each matter type against those criteria, you can make defensible, consistent decisions about optimal delivery models. Here's how that evaluation might look across four example criteria: 

Risk & complexity

• How novel or uncertain is the legal position?
• What's the potential downside if this goes wrong?
• Are there established precedents and patterns we can follow, or does this require original analysis?
• Could a non-lawyer, supported by AI and clear guidance, handle this safely?

Business impact

• How strategically important is this matter to the business?
• What's the financial value or commercial significance at stake?
• Does this require senior stakeholder engagement or relationship management?
• Is this administrative in nature, or does it create competitive advantage?

Volume & predictability

• How often does this type of matter arise?
• How consistent are the patterns and variables across instances?
• Can we create clear rules and decision trees, or does each instance require fresh assessment?
• Is the volume high enough to justify investment in automation or self-service tools?

Judgement requirements

• Does this require negotiation, persuasion, or relationship navigation?
• Are we applying established rules, or developing strategy and novel arguments?
• How much discretion and interpretation is involved?
• Could AI provide reliable guidance, or does this need human judgement throughout?

Designing your delivery model

Once you've evaluated your matter types against the four criteria, you need to design delivery channels that match their characteristics. This requires answering fundamental questions about governance, capability requirements and quality controls for each tier. The following framework provides some key considerations for structuring your delivery model across the three tiers:

Tier 1: Business self-service with AI and legal oversight - key design questions

• What are the distinctive characteristics of work that is truly routine and suitable for business self-service with AI support?
• Which matter types are most closely aligned to those distinctive characteristics?
• How will you define clear guardrails and escalation triggers for when legal must step in?
• What oversight mechanisms will ensure quality and compliance without creating bottlenecks?
• How will you train business users to use AI tools effectively and safely?
• What processes will you implement to capture and learn from exceptions and edge cases?

Tier 2: Lawyer-led with AI augmentation - key design questions

• What are the distinctive characteristics of work that requires lawyer involvement but could benefit from AI handling groundwork or analysis?
• Which matter types are most closely aligned to those distinctive characteristics?
• How will you decide when AI input is sufficient versus when deeper legal judgement is needed?
• What protocols will you set for integrating AI into lawyer workflows without disrupting quality?
• How will you monitor and review the quality of AI-augmented outputs?
• What training and change management will lawyers need to work effectively with AI tools?
• How will you measure the effectiveness and value creation of this hybrid approach?

Tier 3: Pure human expertise - key design questions

• What are the distinctive characteristics of work that requires senior legal expertise and cannot or should not predominantly use AI?
• Which matter types are most closely aligned to those distinctive characteristics?
• How can AI support, if at all, in the delivery of these matter types?
• How will you ensure these high-stakes matters are resourced appropriately with the right level of expertise (whether internal or external)?
• What governance structures will ensure appropriate escalation and decision-making authority?
• How will you balance the cost of premium expertise with the value and risk profile of each matter?
• How will you capture learnings from these unique matters to inform future delivery decisions?

Adopting these tiers may challenge long-standing expectations of premium service for all matters. That change must be handled thoughtfully to avoid eroding confidence. Success also depends on transparent conversations with business stakeholders, so they understand the drivers behind this model and how it delivers greater efficiency and value without compromising legal risk management.

Building the team for your new delivery model

Getting this right also forces a harder question: is your existing team structured for this new model? As AI reshapes what work goes where, it inevitably reshapes what capabilities you need in-house. You may find you need fewer junior lawyers handling routine matters, but more senior expertise for complex work. You'll almost certainly need new capabilities: legal technologists who can configure and maintain AI tools, operations professionals who can design processes and measure outcomes and data analysts who can extract insights from your delivery metrics.

The good news is that you don't necessarily need to hire all these capabilities. Many organisations already have talented operations, technology and data professionals in other functions who can be deployed, often on a shared or project basis. The question is whether these capabilities need to be embedded permanently or can be accessed as needed.

The right framework doesn't just optimise your current team's ways of working; it reveals what your future team should look like and how to build it.

Designing for the future

The shift to AI-enabled delivery isn't just about efficiency. It's about fundamentally rethinking how legal value is created. When you move routine work to business self-service with AI oversight, accelerate lawyer work with AI augmentation, and reserve pure human expertise for matters that truly require it, you're not just doing the same work faster. You're lowering costs, improving risk management and getting higher satisfaction from both lawyers and the business. But the real prize is strategic: repositioning legal from a necessary cost centre to a genuine business enabler.

The framework outlined here - segmenting work through systematic assessment, designing a tiered delivery model, and implementing clear processes and governance - provides the structure to make this shift deliberately rather than reactively. But the framework alone isn't enough. You also need the courage to challenge how things have always been done, the discipline to measure what matters and the willingness to evolve your team's capabilities and structure to match your new delivery model.

The question therefore isn't whether to change your delivery model. It's whether you'll lead that change or be forced into it by business pressure. What will you choose?

• Redesign their service delivery models to align them with their strategy
• Document their service catalogue and prioritisation framework to deliver greater value
• Build scalable legal teams by shaping fit-for-purpose teams in the age of AI
• Identify and prioritise AI use cases that deliver measurable value
• And much more.

Let us help you decide the right path, protect what matters, and evolve your function to deliver lasting impact. Contact us to learn more.