WIRED reported on January 9, 2024, that Google has added a new feature to Gmail called "AI Inbox", which provides automatic summaries and action suggestions at the inbox level. When users open the inbox, an AI banner condenses long email threads or multiple unread messages: it lists key points, extracts to‑dos (such as times, locations, tasks) and offers several reply drafts in different styles.
According to the report, the AI Inbox is powered by Google’s Gemini model. The interface does not replace whole passages with a single sentence; instead it displays a compact card at the top of the message list: a one‑ or two‑line summary sentence followed by bullet points of "Key facts" and "Next steps." Users can click a suggestion to insert it into a reply draft with one click, or choose different tones (polite, direct, concise) to have the system rewrite the candidate reply.
The WIRED article describes operational details: when the system detects multiple related messages it automatically groups them into a single topic card, and the card provides "three reply styles" and "actionable suggestions" (such as scheduling a meeting, confirming information, or requesting more details). If users need a more detailed summary, they can expand the card to view the timeline and key people identified by the model.
Google positions these features as an extension of embedding AI into everyday communication: not just a drafting assistant, but an attempt to reduce the amount of content users must personally scan at the inbox level. WIRED notes that the features will be available to both personal Gmail users and Workspace enterprise customers, but rollout timing and country/region coverage will be phased.
On privacy and data use, the report states that Google says these AI features are designed in line with existing privacy policies and emphasizes that Workspace customer data will not be used by default to train models (users and organizations can configure settings). The article also cautions that any design that sends email text to large models raises questions about data flows and retention policies, and enterprise customers should assess compliance before enabling the features.
In WIRED’s hands‑on experience, the AI‑generated summaries usually capture meeting times and task assignments, but sometimes fall short in recognizing the tone and implied meaning of multi‑party replies and require user adjustment. The system’s "to‑do" prompts are particularly valuable for busy inboxes, surfacing messages that require a response out of the information noise.
Google’s move represents a more proactive integration of the Gemini model into everyday office workflows — expanding from writing assistance to the "reading and decision" stages. WIRED suggests this change will affect how people process email: greater reliance on models for filtering and summarizing, while still requiring users to review and vet generated content.
▶️ Like this article? Creating content is hard!
Follow our YouTube channel:
AI Rosetta
Your likes, subscriptions and shares are our greatest support! 🙏
For more AI news visit: https://www.airosetta.com