GitHub Copilot

GitHub Copilot is an AI programming assistant developed by GitHub, a subsidiary of Microsoft, that integrates into code editors to suggest and generate code. First released as a technical preview in 2021 and made generally available in 2022, it was among the earliest widely adopted developer tools built on large language models. It began as an autocomplete engine offering inline suggestions and has since expanded into a broader assistant that can converse about code, review changes, and act autonomously on multi-step tasks.

From Autocomplete to Agent

In its original form, Copilot analysed the code and comments around a developer's cursor and proposed completions ranging from a single line to an entire function. This inline completion remains a core feature, but the product has grown substantially. Copilot Chat lets developers ask questions in natural language about a codebase, request explanations, generate tests, or ask for fixes. A code review capability can comment on pull requests and suggest improvements.

The most significant shift is agent mode, which became available in 2025. Rather than only responding to a prompt, the coding agent can plan and carry out a task across many files: deciding which files to change, editing them, running terminal commands, observing errors, and iterating until the task is complete. This positions Copilot alongside a new generation of autonomous coding tools and reflects a wider move from suggestion toward delegation in software development.

Models and Editor Support

Copilot originally ran on a model derived from OpenAI's Codex, itself descended from the GPT family. Over time GitHub added model selection, so that on paid plans developers can choose among frontier models from multiple providers, including OpenAI's GPT series, Anthropic's Claude models, and Google's Gemini, picking whichever suits a given task. Copilot integrates with major development environments including Visual Studio Code, Visual Studio, the JetBrains family of IDEs, and others, and offers a command-line interface for terminal workflows.

Plans and Pricing

GitHub offers Copilot across several tiers. A free tier provides a limited monthly allowance of completions and chat interactions suitable for casual use. Paid individual plans, Copilot Pro and Pro+, unlock higher limits and access to the full range of frontier models, while Business and Enterprise plans add organisation-wide administration, policy controls, and security features. In 2026 GitHub moved parts of its pricing toward token-based billing using credits, a change that raised costs for heavy users of agentic workflows and drew criticism from some developers. Pricing and limits continue to evolve as the underlying model economics change.

Impact and Debate

Copilot is credited with measurable productivity gains, with GitHub's own studies reporting faster task completion for developers using the tool. It has also intensified debates central to AI-assisted coding: the risk of introducing subtle bugs or insecure code, the need for developers to review generated output critically rather than accept it blindly, and unresolved questions about the copyright status of code produced by models trained on public repositories. A class-action lawsuit brought against GitHub, Microsoft, and OpenAI over the use of open-source code in training was largely dismissed, though the underlying legal questions remain contested. Copilot has nonetheless become a reference point for the category of AI pair programmers and has spurred a competitive market of rival assistants.

GitHub Copilot is widely used across Malaysia's software and digital services sector, from multinational development centres in Cyberjaya and the Klang Valley to startups and freelancers. For a country positioning itself as a regional technology hub under the MyDigital Blueprint and MSC Malaysia, tools that raise developer productivity have direct economic relevance, helping smaller teams ship more with constrained talent.

Adoption also intersects with the national skills agenda. MDEC and HRD Corp fund training in software engineering and increasingly in AI-assisted development, and technical institutions and coding bootcamps have begun teaching students to use assistants like Copilot responsibly, emphasising code review and security rather than uncritical acceptance of suggestions. This matters because generated code can carry vulnerabilities, a concern relevant to CyberSecurity Malaysia and to firms building software for regulated clients in banking and government.

Data governance is a consideration for Malaysian enterprises and public agencies evaluating Copilot. Organisations handling personal data under the PDPA, or sensitive government code, must weigh what context is sent to cloud-hosted models and prefer business or enterprise arrangements that offer stronger controls over data handling. As Malaysia pursues its goal of becoming an AI-ready nation with a modern GovTech platform, AI coding assistants are becoming part of the everyday toolkit for the developers who build that infrastructure.

1. GitHub. (2025). GitHub Copilot documentation and plans. github.com/features/copilot.
2. GitHub Docs. (2026). Models and pricing for GitHub Copilot. docs.github.com.
3. Chen, M., et al. (2021). Evaluating Large Language Models Trained on Code. arXiv:2107.03374.