Google has finally set a definitive date for the retirement of Google Assistant on Android phones. According to an email sent to users, Assistant will begin shutting down on September 4, with the transition to Gemini expected to take several weeks. The move ends a long run for the voice assistant that first debuted in 2016 and signals a major turning point in Google's AI strategy.

Key facts at a glance

• Google Assistant on Android devices will begin shutting down on September 4.
• The rollout of the shutdown may take several weeks to complete.
• Users of smartwatches, headphones, and Android Auto will also be moved to Gemini.
• Android TV, set-top boxes, and cars with Google built-in are not yet on the shutdown list.
• Devices that do not meet Gemini's minimum specifications or are in unsupported regions may retain Assistant.
• Users can disable voice control entirely if they do not want Gemini.

Why Google is retiring Assistant

Google's original plan was to retire Assistant in late 2025, but the company delayed the process as it worked to bring Gemini up to standard. Now that Google considers Gemini ready for widespread use, the old assistant is finally being put out to pasture. The email sent to users describes the change as part of Google's broader effort to consolidate its AI products under the Gemini brand.

Assistant has been available on Android devices for nearly a decade. It grew out of Google Now and became a central feature of Android phones, smart speakers, and thousands of third-party devices. Over time, however, Google began shifting resources toward Gemini, a large language model-based assistant that can hold conversations, generate text, and use tools. The company has been pushing Gemini as the future of its AI-powered services, and today's announcement is the clearest sign yet that Assistant's days are numbered.

What the shutdown means for users

When the shutdown begins on September 4, users who are still relying on Assistant will be prompted to switch to Gemini. For those who have already made the change, nothing will change. For the rest, the familiar "Hey Google" voice commands will now be processed by Gemini instead of the older Assistant software.

The transition will not be limited to phones. Smartwatches, wireless headphones, and cars with Android Auto are also affected because they depend on the phone's assistant to handle voice requests. That means users who ask their watch to start a workout or tell their headphones to send a text will be interacting with Gemini even if they never installed the app on their phone.

Google's smart home devices have largely already transitioned to Gemini, so the main impact is on mobile and wearable ecosystems.

What still works with Assistant

Assistant will not disappear from every product on September 4. The email notification does not mention Android-powered TVs, set-top boxes, or vehicles with Google built-in. These platforms are expected to transition to Gemini in the future, but Google has not committed to a date yet.

There is also some good news for people living in regions where Gemini is not available or using older phones that do not meet Gemini's minimum requirements. In those cases, Assistant may remain active for a while longer. However, most users will lose access within the next month or two, according to the email.

What replaced Assistant

Gemini is a fundamentally different kind of assistant. Instead of handling predefined commands and simple queries, it is powered by a large language model capable of generating responses, reasoning about context, and chaining multiple tasks together. It can also call external tools to complete actions that were previously hardcoded into Assistant.

For example, setting a timer or turning on a smart light used to be handled by a specific module in Assistant. With Gemini, the system must understand the request, decide which tool to use, and then call that tool. This approach is more flexible in theory, but it also introduces new problems.

Gemini sometimes goes off on a tangent or forgets that it has access to a particular tool. Users have reported situations where Gemini answers a question instead of performing an action, or where it apologizes and says it cannot help when the right tool was available. This can be frustrating when all you want is a quick, reliable voice interaction.

The tradeoff is that Gemini can do much more than Assistant ever could. It can draft emails, summarize documents, recognize objects through the camera, and hold natural conversations. It is not just a voice command interface; it is a full AI companion embedded into the operating system.

The history of Google Assistant

Google Assistant launched in May 2016 as part of Google Allo and was later integrated into Google Home and the Pixel smartphone line. It was an evolution of Google Now, which had already introduced predictive cards and hands-free search. Assistant was designed to be conversational, and for a time it was considered one of the best voice assistants on the market.

Over the years, Assistant expanded to more than a billion devices. It powered Android phones, Google Nest speakers, Chromebooks, smart displays, and even washing machines and light bulbs. Developers could build actions for Assistant, and device makers could embed it into their products through the Google Assistant SDK.

But the rise of generative AI changed the landscape. OpenAI's ChatGPT showed that conversational AI could be much more powerful than a rule-based assistant, and Google soon began working on a response. Gemini, originally launched as Bard, was Google's answer to ChatGPT. The company gradually integrated Gemini into Search, Workspace, Android, and now the assistant role that was once reserved for Assistant.

What users should do now

For most Android users, the only real choice is whether to switch to Gemini or disable voice control. If you want to keep using voice commands, setting up Gemini is straightforward. You can download the Gemini app from the Play Store or enable it through the Google app settings. Once active, Gemini will handle voice requests from the phone, watch, and other connected devices.

If you prefer not to use Gemini, you can turn off voice controls entirely. That means no assistant to set alarms, answer questions, or control smart home devices with voice. It is a drastic step, but it is now the only way to avoid the new AI-powered system.

For users with older devices, the transition may not be immediate. Google has not published a full list of supported devices for the Assistant shutdown, but Gemini has been gradually expanding its device compatibility. If your phone lacks the required RAM, processor, or software version, Assistant may continue working until you upgrade.

The future of Google's AI assistant

The shutdown of Assistant is part of a larger shift inside Google. The company has been consolidating its AI products under the Gemini name and reducing investments in older systems. In the past year, Google has replaced many of its standalone AI features with Gemini-branded services, and the assistant is the latest example.

It is not hard to understand why Google wants to make this move. Maintaining two assistants is expensive, and Google would rather focus its machine learning research and infrastructure on a single platform. Gemini can also generate revenue through premium subscriptions, while Assistant did not have a direct monetization model.

But the transition is not without risks. Many users have grown accustomed to Assistant's predictable behavior. They know exactly how it will respond to simple commands, and they rely on it for tasks like checking the weather, navigating home, or playing music. Gemini's conversational style can feel unpredictable by comparison, and its tool-calling capabilities are still not perfect.

Still, Google is committed to this path. The company has invested heavily in Gemini and has made it the default assistant on Pixel devices. Over time, Gemini is likely to improve, and the rough edges that users are experiencing now will be smoothed out with software updates. The question is how many users will be willing to put up with the transition period.

The September 4 date may be the beginning of the end for Assistant, but it is not the end of voice assistants as a whole. It is the beginning of a new era where voice interaction is driven by large language models rather than predefined rules. Whether that is better or worse will depend on how quickly Google can make Gemini reliable enough for everyday use.

Source: Ars Technica News