Google Gives Gemini Its Own Email Address and a Job: What the Workplace Agent Means for Alphabet Stock

Alphabet's Google unveiled a Gemini agent on Thursday that is designed to work like a colleague, not a chatbot. It operates inside Gmail, Drive, Docs and Calendar, can have its own corporate identity, and can pursue a goal across several days. The launch widens Google's fight with Microsoft, OpenAI and Anthropic for corporate customers, the business Alphabet needs to justify its hundreds of billions of dollars in AI infrastructure spending.

What Google actually announced

Google Cloud CEO Thomas Kurian introduced the agent at the company's Gemini at Work event. Google says it can answer questions, create content, write and run code, generate images and media, schedule tasks and use a person's data and tools, including software that is not Google's. Kurian said it can work as a personal assistant or as a team member, such as a project manager for a group or an analyst for a finance department.

• Its own identity. The agent can get its own Workspace account, email address, calendar and Drive storage. Its actions show up in an audit trail attributed to the agent, not to a person.
• Multi-day work. It can pursue a goal over days, keep context so users do not have to re-brief it, and spin up temporary sub-agents for parts of a job.
• Model choice. It picks the best model for each job. Google's own models and Anthropic's Claude are supported today, with other models planned.
• Cost and security controls. Real-time spending caps can pause an agent when a project's budget runs out, and Google cites governance and audit features for corporate buyers.
• Reach beyond Google. It works in Microsoft 365 and Slack and connects to tools such as Salesforce, ServiceNow, Jira and Snowflake, plus any Model Context Protocol server.

Google also plans industry-specific Gemini products. Financial services and legal are in preview, with government, healthcare and retail listed as coming soon.

How Alphabet stock reacted

Alphabet shares rose as much as about 1% on Thursday, outperforming a broader pullback in tech, according to market reports during the session. Intraday quotes varied by source and time of day, so treat the move as approximate until the closing figure is published. In the morning, GOOG was up about 0.7% near $350.

Analysts are broadly positive. The consensus rating is Strong Buy with an average price target of $438.75. TD Cowen raised its Buy-rated target to $485 on Thursday, per Benzinga. Alphabet is expected to report earnings around Oct. 28, which will be the next chance for the company to show whether enterprise AI is turning into revenue.

The enterprise agent race

Google is not alone. Over the past few weeks, every major rival has launched its own version of an always-on work agent.

One detail stands out: Google's agent already supports Anthropic's models, which signals a multi-model approach rather than an attempt to win only on Google's own model.

What is still unknown

The announcement leaves several investor-relevant questions open. Google has not published a separate price for the agent or said whether it is included in Gemini Enterprise subscriptions. There is no general availability date, and it is unclear whether an agent's Workspace account counts as a billed seat. Google has not provided independent benchmarks of how reliable the agent is on long, cross-app tasks.

The model behind it: Gemini 4 Argon

The agent launch comes as Google rolls out its next flagship model, Gemini 4 Argon, in phases. Google introduced it on Sept. 30 and has begun sharing it with trusted cybersecurity partners first. It has not given a public release timeline. In Google's own release, Argon beat rival models on several benchmarks but trailed on two of four coding benchmarks, according to Reuters. Coding is the area where employees have reportedly voiced concern, so the broader release is worth watching.