GitHub opens Copilot to Australian enterprise users with new credit model

A new credit-based model

GitHub has officially opened all Copilot features to its enterprise customers in Australia. This update removes the requirement for a separate Copilot license for these organizations. Instead, enterprise customers will now pay for AI credits to power their development work.

This shift allows administrators within a company to decide exactly who in their organization can access the tools. By moving to a credit system, the platform aims to align costs with actual usage. This model supports a broader update to the platform's infrastructure as it adapts to more intense AI activity.

The move comes as GitHub reports a significant increase in developer activity across the Australian market. The company is positioning itself to support a landscape where AI is no longer just a side tool but a core part of the workflow. As reported by CFOtech Australia, this transition reflects the rising demand for AI-driven development capabilities.

Scaling for AI agents

To handle the heavier demands of AI agents, GitHub has significantly expanded its physical infrastructure. The company added 4 million CPU cores and 160 petabytes of storage this year. These upgrades are necessary to manage the high volume of data and processing power required by modern software tools.

Beyond hardware, GitHub is reworking its Git architecture to manage a much higher rate of code pushes. This structural change is designed to keep the platform fast and reliable as more developers use automated systems. The goal is to ensure the backbone of the software world can handle the rapid pace of AI-generated output.

The data shows that the shift toward automation is already happening at scale. By September, more than 40% of pull requests on GitHub began with an agent, a sharp increase from 17% in March. This means AI is increasingly taking the first steps in the software development lifecycle.

New tools for complex tasks

GitHub is introducing several new product features to help developers manage these AI agents more effectively. Copilot Fusion is now generally available, allowing the system to select which AI model to use for a specific task. It can even switch between different models or assign different stages of a project to different models.

The platform also launched Copilot Projects, which provides a persistent workspace for agent-based work across multiple repositories. This allows teams to maintain context when working on large, complex systems that span many different folders. Additionally, Copilot Automations now allow workflows to learn from previous outcomes to improve over time.

For those needing more precision, a new Copilot Code Review Max option has been added for more detailed reviews. These tools are designed to help developers move from writing every line of code to orchestrating fleets of agents. The focus is shifting toward shipping completed work, such as bug fixes and security updates, rather than just producing more code.

Growth in the Australian market

Australia has emerged as a fast-growing market for GitHub, with over 2.57 million developers building on the platform by the end of the second quarter. This represents an 8.6% increase from the previous quarter and a 30.0% jump from the previous year. Code uploads in the country also surged by 110.8% year over year.

The diversity of languages used in Australia is also evolving. While JavaScript remains the most popular language for unique developers, infrastructure-related languages are seeing a rise. TypeScript moved from 11th place in 2020 to fourth, while Dockerfile and Makefile now rank in the top six.

The industry breakdown for Australian public repositories shows that Information is the leading sector. Other major sectors include Educational Services, Finance and Insurance, and Retail Trade. This growth indicates that AI-assisted development is spreading into various industries beyond traditional software engineering.

The future of software work

The leadership at Microsoft suggests that the role of the developer is undergoing a fundamental transformation. Jay Parikh, Executive Vice President of CoreAI at Microsoft, noted that the goal of development remains the same: shipping completed software work. However, the method is changing from manual coding to orchestrating AI assistants and fleets of agents.

Parikh emphasized that the infrastructure must be fast and reliable to support this new volume and shape of development. He believes that the software written to date will amount to less than 1% of the software written in the next five years. This suggests a massive expansion in the amount of software that can be created once the bottlenecks of manual coding are removed.

The opportunity for the industry lies in solving problems that were previously impossible to tackle due to complexity. By using AI to handle the heavy lifting of code generation, developers can focus on higher-level problem solving. GitHub is building the foundation to make this shift possible for enterprises worldwide.