Big Tech companies have been rushing to design and produce new “generative” AI products since OpenAI unveiled ChatGPT last November. But the question of how the companies would get people to use — and pay for — the expensive new technology has swirled around the industry for months. Google, Amazon and Apple all have millions of customers already speaking to the companies’ existing voice assistants to set alarms, check the weather and make notes for them, representing a ready-made group of consumers to test out the new AI chatbots on.

But the tech is still new, and the companies are actively working out bugs, despite launching the tools to millions of people already. In September, Google integrated Bard into Gmail, YouTube and Google Docs, but the bot was quickly found to make mistakes, such as making up emails that didn’t exist when asked to summarize important emails from a user’s inbox. When Amazon showed its new “let’s chat” conversational mode for its Alexa speakers onstage at a company event, the bot spoke with long, awkward pauses.

In Wednesday’s announcement, which came during Google’s “Made by Google” hardware event where it announced two new Pixel phones, Google said that the product, called Assistant with Bard, was still an “early experiment” and that it would be available only to a test group at first. Other users will get access sometime in the “next few months,” but only on mobile devices.

A spokesperson for the company wouldn’t comment on when the chatbot might come to Google’s home devices, which compete directly with Amazon’s Alexa.

Adding the new chatbot AI to Assistant would make it more “intuitive, intelligent and personalized,” Sissie Hsiao, vice president of Assistant and Bard, said in a blog post.