New ways to create personalized images in the Gemini app

Ignoring these advancements leaves users tethered to inefficient, manual prompt-writing workflows that fail to capture individual aesthetic preferences. Adopting this integration allows for the rapid creation of highly relevant visual content, such as generating custom marketing assets featuring specific team members or brand-aligned styles. However, reliance on automated context carries the risk of occasional inaccuracies in likeness or style, necessitating manual refinement by the user.

Group Product Managers Animish Sivaramakrishnan and David Sharon have integrated Personal Intelligence into the Gemini app to reduce the friction of manual image prompting. The system now pulls context directly from connected Google Photos labels and user preferences to ground outputs in personal data. While previous workflows required tedious manual uploads, the new method automates context retrieval through Nano Banana 2. A current limitation remains the occasional need for iterative feedback when the AI fails to select the intended reference photo on the first attempt.

This week

• Verify current Google AI subscription tiers for team accounts to ensure access to the rollout.
• Review existing Google Photos organization and labels to ensure accurate metadata for AI referencing.

This quarter

• Establish internal guidelines for using personalized AI imagery in non-sensitive communications.
• Test the refinement workflow to determine the average number of iterations required to achieve brand-consistent outputs.

This year

• Evaluate the integration of personal intelligence models into broader creative workflows to reduce reliance on stock photography.
• Monitor updates to privacy controls to ensure alignment with evolving data security policies.