Google Gemini free AI restores old photos and launches Spark agent

Google’s Gemini AI now offers a free photo‑restoration tool that can repair, sharpen and colorize old family pictures from a single English prompt, while the same platform rolls out the Gemini Spark agent for paid subscribers that can execute tasks across Google Workspace and third‑party services [1][2].

Free photo restoration beats competitors

Komando reports that the Gemini app can take a cracked, faded or blurry family photo, apply a single prompt such as “Restore and colorize this photo,” and return a repaired image in seconds, preserving the original faces better than ChatGPT’s generative approach, which often reimagines subjects [1]. The tool works without Photoshop skills and is accessible through the Gemini website or mobile app. For severely damaged images, Google’s newer “Nano Banana Pro” model can rebuild rough damage, extending the capability beyond simple brightening or background removal.

Gemini Spark expands AI from chat to action

Google announced Gemini Spark as an “active partner” that moves beyond answering questions to performing tasks on a user’s behalf. Integrated with Gmail, Docs, Slides and Calendar, the agent can scan inboxes for updates, synthesize meeting notes and even place orders through Instacart or reservations via OpenTable [2]. The feature is being tested this week and will be rolled out to AI Ultra subscribers—who pay $100 a month—for full access next week. Experts cite the deep integration with Google’s ecosystem as a potential edge over rivals, noting that the agent’s access to personal data could boost its “personal intelligence” but also raise privacy concerns [2][3].

Privacy considerations and user control

Google asserts that when Gemini processes data within Workspace apps, it does not retain the content for training its foundational models, aiming to protect user privacy [3]. Nonetheless, the expanded data access required for task execution—such as reading emails or linking to external services—has prompted caution from analysts, who recommend users start with low‑risk tasks before granting higher‑level permissions [2].

What to watch

• Gemini Spark rollout – monitor the start of AI Ultra subscriber access next week and any subsequent pricing adjustments.
• Privacy policy updates – watch for Google statements on data handling as the agent gains deeper integration with Gmail and third‑party services.
• Competitive response – track announcements from OpenAI, Microsoft and other AI vendors on comparable agentic features.

Google’s free Gemini photo tool democratizes image restoration for everyday users, while the paid Gemini Spark agent signals a shift toward AI‑driven personal assistants that could reshape how consumers interact with their own data and external services. The balance between convenience and privacy will determine how broadly the new capabilities are adopted.