Microsoft has released Visual Studio Code 1.97, pushing GitHub Copilot deeper into the development process with new features aimed at enhancing productivity and security.
This latest update gives Copilot the ability to predict edits, automatically apply changes, and integrates stricter security measures to protect developers from malicious extensions. The addition of GitHub Copilot’s Next Edit Suggestions (NES) brings a more proactive approach to code editing, marking a significant leap from traditional autocomplete features.
The update not only strengthens Copilot’s role in the editing process but also introduces a new extension security system, which mandates that developers verify extension publishers before installation.
These features aim to increase efficiency and safety for VS Code users, continuing the trend of transforming the developer environment through AI-assisted tools.
Furthermore, the Command Palette, a core element in VS Code’s interface, can now be repositioned, providing users with more control over their workspace layout.
Copilot Moves From a Code Finisher to a Code Editor
With the introduction of Next Edit Suggestions (NES), GitHub Copilot is evolving from a simple code completion tool into a more advanced AI-driven assistant. Instead of waiting for a developer to finish writing code, NES anticipates the next change that might be needed based on the developer’s current actions.
For example, if a developer renames a variable, Copilot not only suggests completing that line but also identifies and proposes changes to all instances of that variable elsewhere in the file.
According to Microsoft’s official release notes, “Based on the edits you’re making, Copilot NES both predicts the location of the next edit you’ll want to make and what that edit should be.”
This capability means developers can now rely on Copilot to handle much of the repetitive work involved in editing code, making refactoring and debugging faster and less error-prone.
This represents a major shift for GitHub Copilot. The feature allows it to take a more proactive role in the development process, rather than simply responding to direct input.
The update is in line with Microsoft’s broader vision for AI-powered development tools, as evidenced by earlier features like AI-assisted code reviews in Copilot. While Next Edit Suggestions is still in preview mode, it showcases the potential for AI to take on an even greater role in shaping code, moving beyond the traditional auto-completion tool it once was.
Auto-Accepting AI Edits: More Efficiency or More Risk?
Along with NES, another feature introduced in VS Code 1.97 is the ability to automatically accept edits proposed by Copilot after a configurable delay. With the chat.editing.autoAcceptDelay setting, developers can opt to have Copilot make changes without manual intervention.
Once a suggested edit has been displayed for a certain amount of time, the change will be automatically accepted unless the developer intervenes.
As Microsoft explains, “The countdown stops when you interact with the accept button or start reviewing changes.”
This feature promises to streamline the editing process, particularly for routine tasks that don’t require much human oversight. However, it also raises concerns about accuracy and potential for errors, especially in complex codebases where even small changes can have unintended consequences.
The feature is disabled by default, allowing users to decide whether they want to incorporate this automated functionality into their workflow.
While this feature can save developers time, it also brings forth a critical question: how much autonomy should AI have in the coding process?
This is especially important when considering the risk of unchecked errors, which may be harder to spot if AI is making decisions without human review. Thus, Copilot’s auto-accept feature offers an opportunity to increase productivity but with the potential tradeoff of losing some level of human control.
Expanded AI Model Support: OpenAI and Google Join the Mix
Along with the enhancements to Copilot’s editing features, Microsoft has also expanded the number of AI models available for use within VS Code. Developers now have access to OpenAI’s o3-mini, which offers improved reasoning capabilities, and Google’s Gemini 2.0 Flash, a model optimized for faster response times.
By integrating these models, VS Code now provides more flexibility in how developers choose to interact with AI, allowing for tailored suggestions depending on the specific needs of the task at hand.
The new models are designed to bring more nuanced and adaptable AI support, offering developers different options based on the complexity of the task. For instance, o3-mini may excel in scenarios that require detailed reasoning, while Gemini 2.0 Flash can be beneficial for tasks that prioritize speed and efficiency.
The introduction of these models is part of a growing effort to make AI-powered coding assistants more adaptable. As noted by other AI coding assistants in the market, such as Cursor, which uses its own fork of VS Code, this shift allows developers to choose AI tools that best match their workflow, whether they need reasoning power, speed, or other specialized abilities.
VS Code Extension Security: New Publisher Trust System
Security has been a growing concern in the developer community, particularly regarding third-party extensions in VS Code. With the release of version 1.97, Microsoft has implemented a new publisher trust system to help prevent malicious extensions from compromising users’ development environments.
From now on, developers will be required to approve extension publishers before installing their extensions, adding an additional layer of protection against potentially harmful software. Installing extensions via the command line will bypass this verification process.
This means that while developers installing extensions through the VS Code interface will be prompted to approve the publisher, those using the command-line interface (CLI) will need to manually manage extension trust. This measure is a response to incidents where unverified extensions have been used to inject malicious code into development environments.
By requiring developers to trust the publisher before installation, Microsoft aims to protect the larger VS Code ecosystem and ensure that users are more conscious of where their extensions are coming from.
However, this security feature is not a complete solution—developers must still exercise caution when choosing which extensions to use, especially given the number of third-party tools available in the VS Code Marketplace.
Repositioning the Command Palette for Better UI Control
VS Code 1.97 also introduces a simple yet impactful feature: the ability to reposition the Command Palette and other quick input elements. Previously, the Command Palette was fixed at the top of the window, which meant users had little flexibility in customizing their workspace layout.
Now, developers can drag and drop the Command Palette to the center or any other position within the editor window, allowing for more flexibility in their workspace setup.
This gives developers more control over their work environment, which is particularly useful for those who rely heavily on keyboard-driven navigation. In fact, many developers prefer to keep the Command Palette in a specific location, whether it’s at the center or somewhere else within the UI.
As Microsoft notes, VS Code will remember the user’s placement of the Command Palette across sessions, ensuring that developers don’t have to reconfigure their workspace each time they reopen the editor.
While this might seem like a small change, it reflects Microsoft’s broader efforts to make VS Code a more adaptable IDE. The move towards a customizable UI is a trend seen across many developer tools, where workflow personalization plays an important role in boosting productivity. This change complements other recent improvements to the editor, such as customizable sidebars and dockable panels, that further help developers tailor their workspace to their needs.
AI-Assisted Code Editing: A New Era for Developers
The introduction of Next Edit Suggestions (NES) in VS Code 1.97 marks a significant step toward the future of AI-assisted software development. GitHub Copilot, now with the ability to predict edits and automate changes, is evolving from a simple tool that suggests completions into a more autonomous assistant.
Rather than waiting for developers to specify changes, Copilot can anticipate the next logical steps, significantly reducing the amount of time spent on repetitive coding tasks like renaming variables or refactoring logic.
Microsoft’s effort to integrate AI more deeply into the coding workflow reflects a growing trend where AI plays an active role in shaping code. This move is not just about improving convenience, but also about enhancing developer productivity. A
s AI continues to mature, it may soon take on more advanced responsibilities, such as refactoring entire codebases or even designing new algorithms. These changes signal that AI’s role in development will likely shift from a passive assistant to an essential partner in the coding process.
While this is just the beginning, the implications are clear: by automating mundane tasks, Copilot is enabling developers to focus on more creative and strategic aspects of coding. This trend is already visible in other areas of software development, such as AI-assisted code reviews (as discussed in earlier articles), where Copilot’s role is expanding to help improve code quality even before it is tested or deployed.
The Road Ahead: Copilot and the Future of AI in Development
Looking ahead, it’s clear that AI-powered development tools like Copilot will only become more integrated into daily coding workflows. Microsoft’s decision to incorporate multiple AI models—including OpenAI’s o3-mini and Google’s Gemini 2.0 Flash—suggests that AI is not only becoming more capable but also more customizable for developers. These models allow developers to choose the right tool based on the specific task at hand, whether that’s reasoning through complex code or generating efficient code quickly.
This flexibility will be crucial as AI continues to take on a larger role in the development process. In the future, Copilot may even be able to handle more complex responsibilities, such as automating entire projects or managing codebases autonomously.
For now, though, features like NES and the ability to auto-accept changes are already significantly improving coding speed and reducing human error. These enhancements align with a larger trend in the software development world, where AI is used to streamline processes, freeing up developers to focus on higher-level tasks.
As AI models continue to evolve, the possibilities for what tools like Copilot can achieve are boundless. While these developments are still in their early stages, it’s clear that AI will play an increasingly vital role in shaping the future of programming. For developers, the challenge will be adapting to this new landscape where human collaboration with AI becomes the standard approach to coding.
In the future, AI-driven development tools may be considered the norm, helping developers move beyond simple code completion to fully automated and AI-enhanced software creation. This shift will have far-reaching implications not just for how developers work, but for how software is built and maintained. With VS Code 1.97, Microsoft is laying the groundwork for a future where AI is an indispensable part of the coding process.
