Google I/O 2025: 5 AI tools that improve app development workflow

Google I/O 2025 revealed a wave of AI-powered tools designed to dramatically simplify software development for builders like us. Given the vast number of announcements, even experienced developers can feel overwhelmed by choosing the right tools. To help navigate this landscape, we've highlighted five standout AI tools that we're most excited about trying — after all, we've been building apps since before Android existed. 

We've organized these tools into three categories based on their stability — battle-tested, promising-but-evolving, and cutting-edge experimental — and added our hands-on perspective about where they shine (and where they stumble).

[Editor’s note: To learn more about how we use AI in development, explore our blog series, “Injecting AI into Software Development.”]

Navigating Google's AI development tools landscape

Today's developers are often stretched thin by repetitive coding tasks, manual testing, debugging challenges, and complex UI design requirements. Adding new tools into existing workflows can initially feel disruptive rather than beneficial. Google's latest AI development tool innovations promise to reduce manual labor, automate routine tasks, and accelerate the entire app-building process. But for executives, the question is: How does this translate to the bottom line?

The top challenges developers face when navigating this AI deluge include the sheer volume of tools, the stability vs. innovation tradeoff (many tools are in preview or experimental stages), the complexity of integrating new AI capabilities into existing workflows, the need for cost management, and the skill gaps required to effectively leverage advanced AI features.

To evaluate which tools to use, we recommend the following: 

• Use tools that are generally available and fully supported by Google.

• Evaluate how easily a tool integrates into your existing tech stack.

• Balance the cost of a tool against the productivity gains it delivers. 

• Prioritize tools with proven security. 

• Align tools with your team's existing skills and expertise.

• Avoid experimental features unless you're prototyping or building proof- of- concepts.

Choosing the right tools is a constant balancing act between innovation and stability. Remember, the goal is to empower your teamdevelopers, not overwhelm them with possibilities — especially considering that Google has a history of introducing betas and concept tools but then never fully releasing them.

With that in mind, let's explore the five AI tools from Google I/O 2025 our Android developers are most excited about, categorized by their level of stability:

Stable and battle-tested AI development tools

1. Gemini in Android Studio: Simplified app development

SOURCE: Android Developers Blog

One of the biggest bottlenecks in app development is the time and effort required to code user interfaces and test them repeatedly. Gemini in Android Studio directly addresses this by automating UI code creation and simplifying the testing process. Imagine simply describing an app's function as "a settings screen with a list of options and toggle switches" or showing a mockup image and instantly receiving the required Jetpack Compose code. Gemini also quickly diagnoses crashes, recommending precise fixes, greatly reducing downtime and frustration.

These tools are powered by Google's advanced large language models (LLMs), which have been trained on vast amounts of code and data. This allows them to understand your intent and generate code that is both accurate and efficient. For example, the Journeys for Android Studio feature uses natural language processing (NLP) to execute your test descriptions against different devices and setups, saving you time in manual testing and automated test development. 

Developers can prototype faster, launch features quicker, and spend less time on routine tasks, making this ideal for rapidly growing teams needing consistent quality and speed. This translates to a faster time to market for new features, allowing companies to respond more quickly to changing customer needs and gain a competitive advantage.

Pros:

• Automates UI code creation from text or image prompts.

• Simplifies app testing with natural language test cases.

• Quickly diagnoses crashes and recommends precise fixes, minimizing downtime and improving app stability.

• Streamlines development workflows, reducing manual coding efforts and freeing up developers to focus on more strategic tasks.

Cons:

• May struggle with complex business logic requiring deep domain knowledge, necessitating careful review and validation by experienced developers.

• Limited support for non-UI logic and backend-heavy flows, meaning you'll still need to write code for these aspects of your app.

• May not account for hardware-specific performance nuances in real-time systems, requiring additional testing on real devices.

• Limited support for non-Compose UIs (e.g., traditional XML layouts), making it less useful for projects that haven't yet migrated to Compose.

2. Gemini Code Assist: Enhanced productivity across platforms

SOURCE: Google Cloud blog

Many development teams face pressure to quickly deliver complex projects while maintaining high-quality code standards. Gemini Code Assist helps developers by generating code snippets automatically and suggesting improvements in real-time. With robust security features (SOC 1/2/3, ISO 27001/27017/27018 certifications) and seamless integration into popular IDEs like VS Code and JetBrains, it provides immediate productivity gains while ensuring secure and standardized development practices.

Teams can confidently automate repetitive coding tasks, modernize older applications efficiently, and focus more on creative, strategic challenges rather than mundane details. 

Pros:

• Generates code snippets and suggests improvements in real-time, boosting developer productivity.

• Offers robust security features and integrates seamlessly into popular IDEs, ensuring a secure and standardized development environment.

• Automates repetitive coding tasks and modernizes older applications efficiently, reducing development costs and technical debt.

• Helps maintain high-quality code standards, reducing the risk of bugs and security vulnerabilities.

Cons:

• Chat mode may have slow response times due to heavyweight models, which can be frustrating for developers who are used to instant feedback.

• May struggle with agile refactoring and suggest outdated code, requiring careful review and validation by experienced developers.

• Can conflict with other extensions in VS Code, potentially disrupting existing workflows.

• Lacks repo-wide automation compared to GitHub Copilot Workspace, limiting its ability to automate complex, multi-file changes.

• May generate plausible but incorrect code, especially in niche domains, requiring developers to be vigilant and double-check its suggestions. 

Promising-but-evolving public beta AI development tools

3. Jules (asynchronous coding assistant)

SOURCE: Google Labs

Managing code quality, bug fixes, and dependency updates can quickly overwhelm even experienced developers. Jules operates quietly in the background, taking care of these routine yet critical tasks. It securely connects to your existing code repositories, clones the necessary code to a secure Google Cloud virtual machine (VM), understands the full context of your project, and automatically performs tasks such as writing tests or fixing common bugs.

Jules leverages machine learning (ML) algorithms to analyze your codebase and identify potential issues. It then generates code to address these issues, freeing up your developers to focus on more strategic tasks. 

Pros:

• Automates routine tasks like code quality management, bug fixes, and dependency updates.

• Operates in the background, allowing developers to focus on other tasks without being interrupted.

• Connects securely to existing code repositories and clones the necessary code, ensuring data security and privacy.

Cons:

• Currently supports only Python and JavaScript, limiting its applicability to projects using other languages.

• Requires precise prompts and may struggle with vague requests, requiring developers to be clear and specific in their instructions.

• Still in beta; it may have occasional hangs or unapproved plan executions, potentially disrupting workflows.

• Code complexity may be a challenge, and human oversight is still required.

• Lacks cross-platform support and poses potential security risks.

4. Stitch (AI-driven UI design)

SOURCE: Google Labs

Stitch uses generative AI to create UI designs based on your input — from natural language descriptions or simple images — and integrates directly with popular design tools like Figma. It allows designers to quickly explore different design options and find the perfect look for an app. 

It's perfect for teams needing rapid visualization and initial designs for apps, significantly accelerating early-stage prototyping and simplifying collaboration between design and development teams. 

Pros:

• Generates UI designs from natural language descriptions or simple images, making it easy to create initial designs.

• Streamlines workflows by directly integrating with popular design tools like Figma.

• Accelerates early-stage prototyping and simplifies collaboration between design and development teams.

Cons:

• Still in an experimental stage.

• May struggle with complex layouts and produce generic or misaligned elements, requiring manual tweaking to achieve the desired look and feel.

• Limited font/color customization options, restricting design flexibility.

• Generated HTML/CSS may lack semantic structure or responsiveness, requiring developers to refactor the code to ensure accessibility and performance.

Cutting-edge but experimental AI development tools

5. Firebase Studio (rapid prototyping made simple)

SOURCE: Google for Developers

Turning designs into fully functional apps is often slow and resource-intensive. Firebase Studio changes that by instantly converting designs from tools like Figma into fully operational AI-powered applications. It simplifies backend creation and automates much of the initial setup work, letting teams quickly experiment and validate their ideas.

Firebase Studio leverages AI to automate many of the tasks that are typically performed manually, such as creating database schemas and setting up authentication. 

This tool is particularly beneficial for teams looking to rapidly prototype concepts or launch quick-turnaround projects, reducing the barrier from idea to implementation. 

Pros:

• Instantly converts designs from tools like Figma into fully operational AI-powered applications, significantly reducing prototyping time.

• Simplifies backend creation and automates much of the initial setup work, reducing the complexity of app development.

• Allows teams to quickly iterate and validate their ideas, enabling faster innovation.

Cons:

• Still in an experimental stage.

• May struggle with apps requiring advanced AI/ML integration, limiting its applicability to complex projects.

• Limited customization and control over design adjustments, and AI-generated UIs may lack polish.

• Errors in auto-provisioned backends can be hard to trace without coding expertise, potentially leading to security vulnerabilities.

Integrating AI tools for a competitive edge

At ArcTouch, we've seen firsthand how thoughtfully integrating AI tools can transform development. Pairing these advancements with seasoned developers and designers allows for rapid experimentation, faster prototyping, and higher-quality code. We've found that using the right AI development tools not only accelerates our clients' time to market but also frees us to focus on more strategic initiatives, such as adding innovative features, enhancing user experience, and ultimately delivering more successful apps.

The key to leveraging the power of these AI tools is to start with small experiments, then build confidence with them over time. Generally, you might start with the most stable AI tools and gradually experiment with the less mature ones.

The latest AI tools from Google I/O 2025 offer significant opportunities to simplify development, improve app quality, and foster innovation. Companies that embrace these tools and empower their development teams to use them effectively will be the ones that thrive. Don't get left behind — start exploring these AI tools today and unlock the full potential of your development team.

Need help with your Android app project? 

ArcTouch has been creating lovable Android apps — using the top tools and technology — since the dawn of the App Store. Contact us today if you’d like to learn more about our services.