Amazon’s generative AI upgrade for Alexa is now in the hands of over 100,000 users, according to CEO Andy Jassy. That marks real movement for Alexa+, but it is still a small early step compared with the 600 million Alexa devices already in use.
The rollout matters because Alexa+ is not meant to be a simple voice-command refresh. Amazon is trying to turn a familiar digital assistant into something more conversational, more flexible, and eventually more capable of acting across third-party apps on a user’s behalf.
A measured rollout for a much bigger Alexa base
Jassy shared the Alexa+ user figure on Amazon’s earnings call Thursday. The company first unveiled Alexa+ in February and said at the time that it would reach users in waves over the coming months.
That staged approach is important context. Over 100,000 users sounds significant for a new AI product, but the number is far from the 600 million Alexa devices already out there. Amazon is therefore testing the upgraded assistant at a scale that is meaningful, while still far short of its full potential reach.
The early rollout also shows how cautious major technology companies are being with AI assistants. These products are expected to respond naturally, understand context, and handle tasks that older assistants could not manage. But when those systems sit inside consumer devices, the gap between a demo and a dependable everyday feature becomes hard to ignore.
What Alexa+ is supposed to become
Alexa+ is designed to move beyond the older pattern of preset assistant responses. The new assistant should be able to generate original replies as a conversation unfolds, closer to the voice modes in OpenAI’s ChatGPT and Google’s Gemini than to the older Alexa and Siri systems.
Amazon’s broader ambition is agentic behavior. In plain terms, that means Alexa+ should eventually be able to use third-party apps for the user, rather than merely answer a question or trigger a narrow built-in command.
The source article describes several areas that show the intended direction:
- More natural conversations between users and the assistant.
- Original responses generated on the fly.
- Agentic abilities that could let Alexa+ operate third-party apps on a user’s behalf.
- Practical task completion through integrations with other systems.
That vision would make Alexa+ a more active interface for digital life. Instead of waiting for exact phrasing or relying on fixed responses, the assistant is meant to interpret what the user wants and help carry it out.
Important features are still missing
The current version of Alexa+ does not yet include everything Amazon showed when it introduced the product. As The Washington Post reported at launch, Alexa+ lacked some of the key features demoed in February.
According to the source article, the rollout version did not have the ability to use third-party apps like GrubHub. It also did not have the ability to generate a bedtime story for children or brainstorm a gift idea.
It remains unclear when those features will arrive. Jassy said on the call, “We have a lot more functionality that we plan to add in coming months.”
That distinction is central to understanding where Alexa+ stands now. Amazon has begun putting the assistant in front of real users, but the product is still in a build-out phase. The most ambitious capabilities are either incomplete, unavailable, or still waiting for broader release.
The agent problem is accuracy
During his opening comments, Jassy called Alexa+ one of the first action-oriented AI agents for consumers. At the same time, he said the technology is still “primitive” and “inaccurate.”
That admission gets to the core technical challenge. A conversational assistant can be useful when it answers questions well, but an action-oriented agent needs to do more than talk. It needs to complete multi-step tasks correctly, especially when it is interacting with tools, apps, and other systems.
Jassy said most multi-step AI agents currently have a low accuracy rate, between 30% and 60%. Amazon’s goal for Nova Act, the web-browsing agent that powers Alexa+, is 90% accuracy in this domain.
Those numbers show why the rollout is gradual. An assistant that misunderstands a question may disappoint a user. An agent that takes the wrong action can create a much bigger trust problem. For Alexa+ to become a consumer AI agent rather than a conversational demo, reliability has to improve.
Amazon is moving faster than Apple, but both face friction
The source article notes that Amazon’s Alexa+ rollout appears to be progressing faster than Apple’s rollout of its new, LLM-powered Siri. Apple CEO Tim Cook was asked about new Siri delays on Apple’s Thursday earnings call, which happened at the same time as Amazon’s. Cook said the company needed “more time to complete the work.”
Still, the broader pattern is similar. Both Apple and Amazon have reportedly run into snags and delays while trying to add generative AI to legacy digital assistants.
One of the biggest hurdles is getting LLMs to use tools and integrate with other systems. That work is necessary if assistants such as Alexa and Siri are going to complete practical tasks, including setting timers and reading texts. The source article says implementing those integrations has proven harder than expected.
For Amazon, the milestone of over 100,000 Alexa+ users is therefore not the finish line. It is an early public test of whether a familiar voice assistant can become a more capable AI agent while still being reliable enough for everyday use.