Making an app with Xcode may soon become easier than ever with AI, as an Apple presentation highlights the different use cases and capabilities of agentic coding.
While WWDC 2026 saw all eyes on Siri AI, as Apple rolled out features that were supposed to arrive two years ago, the company also unveiled AI-focused developer tools. Game Porting Toolkit 4 includes support for agentic coding, and the same is true for Xcode 27.
In a nearly 90-minute video, recorded at the Steve Jobs Theater, Apple detailed the capabilities of Xcode 27 and its AI integration. In an era where even Apple executives acknowledged people's fears about being replaced by artificial intelligence, Apple sought to portray Xcode's agentic coding capabilities as a powerful extension.
Apple's presentation says that Xcode 27's agentic AI support is meant to give developers meaningful help during the coding process, with AI that understands and thinks in Swift. The company emphasized that Xcode 27 doesn't just offer AI as a slapped-on layer, but that AI is now a core part of the app.
The new Core AI framework lets developers use on-device AI models with ease, delivering strong performance through a modern Swift API. The open-source MLX framework also received upgrades, letting developers experiment and fine-tune AI models.
With a single prompt, Xcode 27 can apply changes across an entire codebase, editing multiple files as needed. Support for third-party AI models, such as those from Anthropic, OpenAI, and Google, is also available.
Developers can start conversations with AI in Xcode 27 right from the toolbar, with an experience that looks somewhat like the new-and-improved Siri AI. Multiple AI conversations can even be maintained simultaneously.
Xcode 27 lets developers have conversations with AI agents. Image Credit: Apple
Apple's AI agent can suggest ideas for apps and app designs based on prompts and provided assets like icons. In short, Apple has effectively embraced vibe coding, even debuting an AI-generated app as part of its Xcode 27 presentation.
Even after an Apple is built, developers can make changes simply by using AI. Backgrounds, effects, animations, features, and translations can be added via conversations with AI agents.
Plugins, in Anthropic's format, can be integrated as well, and developers can also tap into Siri and App intents for their app development needs. This lets apps leverage Siri's understanding of natural language, its web search capabilities, and its Visual Intelligence features.
Siri AI will be able to take actions within third-party applications, like setting timers and alarms, among other things. In a way, the possibilities seem endless.
The AI-focused enhancements in Xcode 27 look as though they will make app development easier than ever. Still, Apple's choice of vibe-coding an entire app on-stage doesn't go too well with Craig Federighi's statements about AI not being a replacement for human interaction.
Overall, it remains to be seen how quickly developers will embrace AI, and just how many vibe-coded apps we'll see in the App Store as a result.
