What CIOs need to know about Microsoft's AI licensing landscape. Microsoft 365 Copilot is a $30/user/month AI add-on that integrates generative AI directly into Word, Excel, PowerPoint, Outlook, and Teams. It requires an existing M365/O365 subscription and is not included in any existing bundle. CIOs must navigate licensing prerequisites, budget for phased adoption, and evaluate ROI before committing to enterprise-wide deployment.
This playbook covers each product's licensing model, budgeting strategies, ROI evaluation frameworks, and forward-looking guidance on how Microsoft's AI licensing is likely to evolve.
1. Microsoft 365 Copilot Licensing Model
Microsoft 365 Copilot is priced at $30/user/month as an add-on to existing M365 E3, E5, Business Standard, or Business Premium subscriptions. Unlike base M365 plans, Copilot is licensed per named user, not per device, and requires active assignment to the specific users who will use it. For a detailed breakdown of all Copilot plan tiers, prerequisites, and commercial structures, see our Microsoft 365 Copilot enterprise licensing guide.
Prerequisites: Users must have an active M365 E3 or E5 (or Business Standard/Premium) subscription, an Entra ID P1 or P2 licence (included in E3/E5), and SharePoint Online and OneDrive for Business enabled. Meeting recordings in Teams must be enabled for Copilot's meeting intelligence to function.
Cost Example: A 500-user organisation deploying Copilot to 100 selected users: 100 users × $30/month × 12 = $36,000/year in Copilot add-on fees on top of existing M365 base licences.
Minimum commitment: Microsoft originally required a 300-seat minimum for M365 Copilot, subsequently reduced to allow smaller deployments. Check current terms at time of purchase — Microsoft has adjusted these thresholds.
2. GitHub Copilot Licensing Model
GitHub Copilot is a separate AI product from M365 Copilot, targeted at software developers. Three tiers:
- GitHub Copilot Individual: $10/user/month (annual plan), $13/user/month (monthly).
- GitHub Copilot Business: $19/user/month. Adds enterprise controls, policy management, and security features.
- GitHub Copilot Enterprise: $39/user/month. Adds Copilot Chat in GitHub.com, repository-aware responses, and custom model fine-tuning.
For enterprise CIOs, GitHub Copilot Business is typically the minimum tier that delivers the security controls, audit logging, and policy management required in corporate environments. Enterprise deployments across 500 or more developers at $19/user generate approximately $114,000/year in additional licensing cost.
3. Dynamics 365 Copilot Licensing
Dynamics 365 Copilot features are embedded at no additional charge within specific Dynamics 365 modules. The pricing model has evolved: core Copilot capabilities for Sales, Service, Finance, and Supply Chain are now included in base Dynamics 365 licences at no extra cost. Role-specific AI agents — such as Copilot for Sales at $50/user/month and Copilot for Service at $50/user/month — remain as separate add-ons that deliver substantially deeper AI integration than the included features. CIOs should audit which features are already included in their Dynamics licences before purchasing additional Copilot add-ons.
4. Azure OpenAI Service Licensing
Azure OpenAI Service operates on a consumption-based model: you pay for tokens processed (input and output), not per user or per month. Pricing varies by model — GPT-4 Turbo, GPT-3.5 Turbo, DALL-E — and by deployment type (global standard vs. provisioned throughput). For predictable enterprise workloads, Provisioned Throughput Units (PTUs) provide reserved capacity at a fixed monthly cost, eliminating per-token variability. PTU commitments start at approximately $2 per hour per PTU and require minimum one-month commitments. For variable or lower-volume workloads, pay-as-you-go token pricing is more cost-efficient.
5. Budgeting & Cost Planning for Copilot Adoption
The total cost of M365 Copilot adoption extends beyond the $30/user/month licence fee. CIOs must account for: the base M365 licence requirement (if not already in place), Entra ID P1/P2 costs (included in E3/E5, additional cost for non-E3/E5 users), change management and training investment (typically 15 to 25 percent of first-year licence cost), IT administration and governance (monitoring adoption, managing assignments, reviewing usage), and integration development if connecting Copilot to internal data sources via Microsoft Graph.
Phase Adoption Strategy: Deploying to 100 high-ROI users in the first year, then expanding based on demonstrated value, is consistently more cost-effective than immediate enterprise-wide deployment. Start with roles where AI assistance in document creation, meeting summarisation, and email management can be measured: executive assistants, consultants, analysts, legal teams, and software developers.
Total Cost of Ownership: For a 500-user phased deployment (100 users year 1, 300 users year 2, 500 users year 3): Year 1 total cost including training and administration approximately $75,000 to $85,000; Year 2 expanding to 300 users approximately $140,000 to $155,000; Year 3 full 500-user deployment approximately $220,000 to $245,000.
Negotiation Strategies: Microsoft EA customers can negotiate Copilot as part of EA renewal discussions. Volume discounts are available at enterprise scale. Microsoft has demonstrated flexibility on minimum commitment thresholds, pricing tiers, and step-up terms for organisations making multi-year commitments.
6. ROI Evaluation Framework
Microsoft's commissioned Forrester study reported 132 to 353 percent ROI for M365 Copilot deployments over three years. These figures require independent scrutiny — ROI is highly dependent on adoption rates, use case selection, and baseline productivity measurement methodology. Our dedicated guide on how to justify the Copilot $30 per user cost provides the role-based modelling framework that passes CFO scrutiny.
Measurable ROI drivers: Meeting summarisation (time saved per meeting × frequency × user value), email drafting (time saved per email × volume × user value), document creation acceleration (time saved per document × frequency), code generation for GitHub Copilot (development velocity improvement × developer hourly cost). The CIOs who achieve the strongest Copilot ROI run a disciplined measurement programme: baseline productivity data before deployment, defined target metrics per use case, 90-day adoption reviews, and a formal business case update at 180 days.
7. Forward-Looking Guidance
Microsoft's AI licensing is evolving rapidly. Several developments CIOs should anticipate:
Copilot agents and autonomous workflows: Microsoft is expanding Copilot into autonomous agent capabilities through Copilot Studio. These agents introduce new licensing dimensions — Copilot Studio is licensed on a consumption basis (messages/month), and as agents become more capable, licensing complexity will increase.
Bundle evolution: Microsoft has bundled Copilot features into higher M365 tiers over time. E5 now includes some Copilot capabilities that previously required separate purchase. Monitor bundle compositions at each EA renewal — the feature set included in your existing licences may have expanded without explicit notification.
Microsoft's AI usage data retention policies: CIOs must ensure that organisational data used by Copilot, including meeting content, email, and document text, is governed under your existing data retention, privacy, and compliance policies. Microsoft's data residency commitments for Copilot features may differ from standard M365 commitments — verify this before deployment in regulated industries.
For CIOs evaluating Copilot adoption, the essential first step is understanding your organisation's baseline M365 licensing posture. Work with your Microsoft licensing partner or independent advisor to map current subscriptions, identify Entra ID gaps, and model phased deployment costs. Request a formal pricing quote with volume discounts before committing to minimum seat purchases. Build a 12-month adoption plan with defined ROI metrics per user segment, starting with high-impact use cases before scaling enterprise-wide.
Microsoft's Microsoft Knowledge Hub contains detailed guides on EA optimisation and licensing best practices. For organisations deploying multiple AI tools beyond Copilot, our GenAI Knowledge Hub provides independent comparison frameworks across GenAI services and procurement strategies.