Google pushes Android Studio into the AI‑first era
The Android development team announced that Android Studio Rabbit 2 has graduated from the experimental canary channel to a full‑blown stable release. While the IDE already sported a sleek UI and faster Gradle builds, this version is centred around artificial‑intelligence assistance – a move that could reshape how UK developers build the next generation of Android apps, from everyday phone utilities to immersive XR experiences.
Bring Your Own Agent (BYOA): Plug‑in any AI coder you prefer
One of the headline features is the Bring Your Own Agent framework. Rather than being locked into a single Google‑provided model, developers can now select from a catalogue of more than 40 AI coding agents, ranging from OpenAI’s Codex to Anthropic’s Claude and Google’s own Antigravity model. The agents are fully integrated with core Android Studio tools – the emulator, the build system, Logcat, and a set of specialised “Android skills”.
To activate an agent, users simply open the Agent pane, head to Settings → Tools → AI → Agents, authenticate the chosen service and start prompting. The flexibility is aimed at three practical scenarios:
- Compliance – organisations bound by enterprise agreements can pick an ACP‑compliant provider.
- Cost optimisation – smaller projects may opt for a lower‑priced model without sacrificing functionality.
- Specialised knowledge – teams can bring in a model that has been fine‑tuned on their own codebase.
For UK developers working on contract apps, this means they can align the AI assistant with the client’s preferred vendor, potentially saving on licence fees while still benefitting from AI‑driven suggestions.
XR emulators finally reach general availability
The XR emulator suite – covering XR glasses, audio glasses and display glasses – is now marked as stable. Previously only accessible in the canary builds, the emulators let developers prototype mixed‑reality experiences directly on a laptop, without needing physical hardware.
Key improvements include:
- Custom 360° backdrops – developers can import their own panoramic photos or stream video from a connected camera, letting the virtual environment mirror a real‑world location.
- Hardware‑level LED indicators for microphone and camera status, making debugging of sensor‑driven apps more transparent.
- A refined Pairing Wizard that links a virtual phone with virtual audio/display glasses faster and with fewer hiccups.
To launch an XR device, open the Device Manager and select the desired emulator type. For UK firms eyeing the burgeoning AR/VR market – think retail visualisers or remote‑training tools – the ability to test on a stable emulator could shorten development cycles and reduce the need for costly hardware trials.
Managing Android‑specific AI skills inside the IDE
Android Studio now ships with a Skills Manager that lets developers discover, install and toggle AI‑driven “skills”. These are essentially pre‑written prompts and workflows that the AI can invoke for common tasks such as:
- Migrating legacy XML layouts to Jetpack Compose
- Upgrading to Android Gradle Plugin (AGP) 9
- Adding adaptive layouts for foldable phones and tablets
- Integrating CameraX or Firebase services
The manager is reachable via the three‑dot menu in the Agent panel, by typing @ in the chat window, or through Settings → Tools → AI → Skills Manager. By curating the skill set, teams can enforce best‑practice patterns across multiple developers – a handy feature for agencies handling several client projects simultaneously.
LeakCanary gets an AI boost
Memory‑leak detection has long been a pain point for Android developers, especially when dealing with complex UI lifecycles. The new release embeds LeakCanary directly into the IDE’s AI assistant. When a leak is detected, the assistant not only points out the offending reference chain but also generates a plain‑English explanation and suggests concrete code changes.
The workflow lives inside the Profiler tab under a dedicated LeakCanary task. Clicking the suggestion opens a diff view where the AI‑generated patch can be applied with a single keystroke. For freelancers and contract developers, this could translate into faster bug‑fix turnaround and fewer hours billed for manual debugging.
Gemini Enterprise: AI with corporate‑grade controls
Large organisations that need strict data governance can now rely on Gemini Enterprise inside Android Studio – a capability that debuted with the Rabbit 1 release but has been expanded for Rabbit 2. The offering focuses on three pillars:
- Zero‑retention privacy – source code and context never leave the organisation’s secure perimeter, and are never used to train external models.
- Audit logging – every agent interaction, generated patch and prompt is recorded, satisfying compliance requirements such as ISO 27001 or GDPR.
- Admin policies – IT admins can centrally whitelist agents, skills and model endpoints, ensuring that only approved AI tools are used across development teams.
For UK enterprises that handle sensitive data – for example, fintech or health‑tech apps – these controls make AI assistance viable without compromising regulatory obligations.
AI‑driven Change Review streamlines code audits
As AI tools accelerate code generation, the bottleneck shifts to reviewing those changes. Android Studio now includes an AI Change Review feature that can be launched from three visual entry points or straight from the Agent chat window.
The process works in two stages:
- Recommended Fixes – the AI scans pending or committed diffs and flags potential issues, offering quick‑fix suggestions.
- Guided Walkthrough – after addressing the fixes, developers can start a walkthrough that groups related changes into logical themes, rather than forcing a line‑by‑line read.
This approach mirrors the way code‑review tools like GitHub’s CodeQL operate, but with the added benefit of being tightly coupled to the Android Studio UI. For teams that rely on peer review for contract compliance, the feature could shave hours off the review cycle.
Getting started and where to learn more
To try the new capabilities, simply download the latest stable build of Android Studio from the official site. The release notes include a quick‑start guide, and the Android developer community is active on LinkedIn, YouTube and X (formerly Twitter) for deeper demos. Google is also streaming a series of talks from the Android Dev Summit – a useful resource for anyone wanting to see the AI tools in action.
Why UK developers should care
While the announcement is technical, its implications are practical for the UK Android ecosystem:
- Faster prototyping – AI agents can scaffold boilerplate code, letting developers focus on unique business logic.
- Cost control – BYOA lets freelancers pick cheaper models, and the Skills Manager enforces reusable patterns across multiple client projects.
- Compliance‑ready AI – Gemini Enterprise addresses the data‑privacy concerns that have held back AI adoption in regulated sectors.
- XR readiness – Stable emulators mean developers can explore the growing market for mixed‑reality apps without upfront hardware spend.
In short, Android Studio Rabbit 2 equips developers with a toolbox that blends AI convenience with enterprise‑grade safeguards – a combination that could influence the next wave of Android apps sold on the UK market.
The information above is based on Google’s official Android Developers blog post dated 28 October 2026.
