Google has launched a new feature that allows Workspace users to access “Gems,” customized versions of its Gemini AI assistant.

This feature is available in the side panel of apps such as Docs, Slides, Sheets, Drive, and Gmail.

Gems enable users to create specialized chatbots for specific tasks like copywriting or exam preparation. Users can also upload files to provide context for assistance.

Google provides pre-made Gems for tasks including text editing, code generation, and creating sales pitches.

While custom Gems cannot be created directly within Workspace apps, users may create them through the Gemini website or the Workspace side panel.

The update, which supports features like @-mentioning and file access, began rolling out on July 2, 2025.

🔗 Source: The Verge

🧠 Food for thought

1️⃣ The battle for workplace integration defines AI assistant competition

Google’s Gems feature reflects the competitive race between major tech companies to embed AI deeply within productivity suites rather than offering standalone AI tools.

Microsoft Copilot charges $20 per user monthly for its integration with Microsoft 365, while Google prices Gemini at a nearly identical $19.99, showing how these companies are directly competing for enterprise AI dollars 1.

Research indicates these platforms are strategically differentiating based on their native ecosystems. Copilot excels in structured document tasks and advanced data analysis within the Microsoft environment, while Gemini leverages its strengths in Google’s ecosystem 2.

This integration strategy addresses a fundamental business challenge. Studies show only 1% of companies consider themselves mature in AI deployment despite significant investments, suggesting that seamless workspace integration like Gems could accelerate adoption 3.

The approach of customizable AI assistants represents a shift from general-purpose AI to tools that align with specific business workflows, potentially addressing the 30% of work hours that McKinsey predicts could be automated by 2030 4.

2️⃣ The AI productivity paradox hinges on effective implementation and training

Google’s emphasis on customizing Gems for specific tasks addresses a critical finding: despite AI’s potential to automate 50% of current work tasks, actual productivity gains depend heavily on workers’ ability to effectively use these tools 5.