Google expands Gemini with Daily Brief digest and persistent Spark agent

Google's three-part Gemini expansion

Google has announced a substantial update to the Gemini app that pushes the assistant beyond answering one-off prompts toward organising information and carrying out work across its connected tools. According to dev.to's coverage of the announcement, the expansion rests on three components: Daily Brief, a personalised morning digest; Gemini Spark, an always-on personal agent; and tighter integration of Gemini Live into the main Gemini experience. Google frames all of this under what it calls Personal Intelligence, a framework for using context from Gmail, Calendar and other Google products to tailor both answers and actions.

Gemini Live merges into the main app

Gemini Live is now built directly into Gemini rather than sitting as a separate mode. Users can move between quick text queries and longer, free-flowing spoken conversations in a single experience, and Google says microphone improvements are meant to support more natural, extended dialogue. The practical aim is to let someone talk through a plan, refine it, and turn the result into actual work instead of restarting the conversation repeatedly.

Daily Brief assembles a morning digest

Daily Brief draws on a user's Gmail, Calendar and Gemini chats to produce a personalised morning overview. Once enabled, it runs in the background and surfaces prioritised items, organisational cues and recommended next steps. For anyone who currently checks an inbox, a calendar and chat threads separately before deciding what matters, the feature is intended to make that first review of the day more structured — though it still leaves verification and decision-making to the person.

Google plans to roll out Daily Brief first to Google AI Plus, Pro and Ultra subscribers in the United States. The announcement reported by dev.to does not give a date for availability beyond the US.

Spark is the agentic core of the update

Gemini Spark is the more ambitious piece. Google describes it as a 24/7 personal AI agent that performs tasks on a user's behalf while staying connected to Workspace tools including Gmail, Docs and Slides. Crucially, Spark is designed to keep working after the user closes the app or locks their device, which separates it from an ordinary chat session that ends with a single response.

A second dev.to report, citing Google's June 2026 Spark updates, adds detail on cross-device orchestration. Google has demonstrated a workflow that starts on a mobile device and completes on a macOS host computer while the user is away. Tasks and reusable skills are set up conversationally, making Gemini Live's voice interface a natural route to configuring longer-running work. Google has also outlined a roadmap of additional skills, automated workflows and third-party connections, but those should be read as plans rather than shipping features.

Availability is phased, and the reports differ

The two dev.to reports describe Spark's rollout slightly differently. One says Spark begins with trusted testers, followed by a beta for Google AI Ultra subscribers, alongside a macOS app update bringing Spark to desktop. The other states Spark is currently available to Google AI Ultra subscribers aged 18 and over in the United States, with a macOS beta and business deployment through the Gemini Enterprise context also referenced. Neither report provides pricing, and no timeline for broader access has been given — so most Gemini users will not have these features yet.

Privacy and control

Google's privacy guidance, as relayed in the coverage, says Gemini in Gmail does not train on personal Gmail content. The company also says access to connected data is restricted and isolated to requested tasks, with users retaining control over what is shared or executed. Even so, teams connecting operational or customer-facing accounts would be wise to define what an assistant may touch and where human review stays mandatory.

Why it matters

The update marks a shift in what a consumer and workplace assistant is expected to do. Rather than responding to a prompt and stopping, Gemini is being built to hold context across conversations, summarise a day from multiple sources, and execute multi-step work in the background across devices. That combination — conversation, personal context and persistent execution — is the direction much of the industry is moving, and Google's Workspace footprint gives it an unusually deep well of context to draw on.

The caveats are equally important. The rollout is tiered, largely US-only for now, and Spark's most ambitious capabilities remain in testing. The sensible near-term use is bounded, low-risk work with a human reviewing the output — drafting from inbox context, summarising commitments, handing off a defined task between devices — rather than delegating consequential work without oversight. Google has set the trajectory; making it reliable day-to-day is still to be proven.