Gemini to allow anyone to build AI Agent on phone soon

Google is testing new upgrades for Gemini. The likely upcoming upgrades include an option to customise the daily brief and manage skills for its always online agent, Spark. The tech company is also experimenting with giving users greater control over custom AI agents from mobile phones. 

An AI agent is an autonomous software system that uses AI to perceive the environment and accordingly make decisions, executing multi-step actions to achieve a goal. Unlike chatbots that rely on text, AI agents can plan users’ workflows, use external tools and solve complex problems. 

As of now, the AI agent option is available only to Google Workspace and Enterprise users and is limited to the web interface. However, it is expected that version 17.45.14 of the Google app on Android might include the option to create and manage agents from your phone or tablet.

With this version, Google appears to be testing a New agent option in Gemini. The option is located in the sidebar, just below the list of recent chats, and is designed to let users create AI agents for specific tasks.

However, unlike Gemini’s web interface—where Workspace and Enterprise users can describe a custom task and create an agent around it—the option currently visible on the mobile version appears to offer only five pre-designed templates:

Family Coordinator: Helps plan household chores, share events with family members, and coordinate tasks.

Tutor:Assists with complex study topics, homework, quiz creation, and tracking learning progress.

Group Organizer: Helps coordinate outings, gatherings, and other activities involving multiple people.

Marketing Agent: Assists with social media posts, newsletters, marketing campaigns, and other creative work aimed at promoting products.

Finance Agent: Helps users track and manage expenses and receive reminders for upcoming bill payments.

It remains unclear whether Gemini on Android will eventually allow users to extensively customize these agents or whether the mobile version will offer more limited functionality compared with the web experience.

Given the largely personal use cases of these agents, Google may not restrict the feature to Workspace or Enterprise users. It could potentially be made available to a wider audience, including users on the free Gemini tier.

For now, however, these possibilities remain speculative. Google has yet to officially announce the feature or clarify how extensively users will be able to customize these agents. Still, the apparent level of development suggests that an official rollout could be on the horizon.