Microsoft launches AI agent with pay-as-you-go pricing

2026-06-16 · Source: News on Artificial Intelligence and Machine Learning · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Microsoft has introduced Copilot Cowork, an AI agent capable of independently performing office tasks like drafting documents, building spreadsheets, and sending emails. Launched on Tuesday, this release marks a significant shift in Microsoft's billing strategy, moving from fixed subscription fees to a pay-as-you-go model for its software, a first in two decades. This change is driven by the high computing power demands of AI systems, with each task billed based on consumption. While a Microsoft 365 Copilot subscription is still required, individual tasks incur separate charges. To manage costs, the service is disabled by default, and companies can set spending caps. Users will also have the option to select different AI models, such as Anthropic's Opus 4.8 and Sonnet 4.6, or GPT 5.5 for "Frontier" tier customers, with a "significantly cheaper" Cowork 1 model planned for everyday tasks. This pricing model mirrors similar moves by GitHub and Anthropic.

Key takeaway

For AI Product Managers evaluating enterprise AI solutions, you should anticipate and plan for usage-based billing models. This shift, exemplified by Microsoft's Copilot Cowork, means your operational costs will directly correlate with AI task execution and model choice. Implement internal spending caps and empower teams to select cost-efficient models like Cowork 1 for routine tasks to manage your budget effectively.

Key insights

AI agent services are shifting to usage-based billing due to high, variable computational costs.

Principles

• AI agent costs necessitate flexible billing.
• Usage-based models allow cost control.
• Model choice impacts operational expense.

Method

Companies implement AI agent services by requiring a base subscription, then billing per task based on computing power consumed, with options for spending caps and model selection.

In practice

• Compare 4,000 documents in hours.
• Synthesize emails for meeting prep.
• Cap employee AI spending.

Topics

• AI Agents
• Usage-Based Billing
• Microsoft Copilot
• Anthropic Models
• Enterprise AI
• Cost Management

Best for: CTO, VP of Engineering/Data, Executive, Director of AI/ML, AI Product Manager, Consultant

Related on AIssential

• The Subsidy Ended: What Tool-Using Agents Actually Cost — GitHub Copilot shifted to usage-based billing, linking costs directly to AI model and token consumption.
• The Next Wave of Enterprise AI — Enterprise AI is transitioning from experimental use to scalable, cost-effective, and customized agentic systems.
• Microsoft Mulls Using DeepSeek for Copilot Cowork — Microsoft considers DeepSeek V4 or other open-source models for Copilot Cowork to reduce AI assistant operational costs.
• Microsoft brings Anthropic's Claude Cowork into Copilot to run tasks across Outlook, Teams, and Excel — Microsoft Copilot integrates Anthropic's Claude Cowork for autonomous task execution across Microsoft 365 applications.
• The AI Subsidy Era is Over — Agentic AI usage is ending the "AI subsidy era," driving a shift to usage-based pricing due to soaring compute costs.
• Copilot usage reveals AI adoption patterns — AI usage reveals human psychology, prompting shifts in management, global governance, and market strategies.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by News on Artificial Intelligence and Machine Learning.