Coming Soon: The PhotoShelter AI Assistant is a conversational agent built into your media library. You can ask it questions in plain language, and it will take action on your behalf — navigating, searching, organizing, and retrieving information — without you needing to click through menus.
Article Contents:
How to access the PhotoShelter AI Assistant
All PhotoShelter for Brands accounts will have access to the PhotoShelter AI Assistant, which can be accessed anywhere in the PhotoShelter admin menu/Library via the left-hand navigation bar:
What the PhotoShelter Assistant Can Do
Navigation
- Navigate directly to a specific collection, gallery, workspace, or asset — just tell it the name.
- Find a collection or gallery by partial name if you don't remember the exact title.
Search
- Search your library by keyword, metadata fields (IPTC/XMP), upload date, photo date, file name, rating, orientation, media type, and more.
- Perform AI Visual Search (AIVS) — find visually similar images or videos using AI, either by describing what you're looking for or by anchoring to an existing file.
- Scope any search to a specific gallery or collection.
- Show search results in the library view, or return structured results programmatically.
Library Organization
- Create new collections (folders) or galleries, optionally nested inside a parent collection.
- Rename or update the description of a collection or gallery.
- Move collections and galleries to a new location in the library.
- Copy media files from one gallery to another.
- Copy media from a shared organization's library into your own galleries.
File & Metadata Details
- Fetch details about any media file: name, format, dimensions, file size, rating, flag status, and release status.
- Retrieve IPTC/XMP metadata fields: keywords, description/caption, headline, title, author, copyright, credit, source, city, state, country, usage rights, and date created.
- Retrieve EXIF capture data: camera make/model, lens, ISO, aperture, shutter speed, focal length, and GPS.
- Retrieve AI-generated tags from models like ObjectID, PeopleID, PlayerID, and LogoID.
- Retrieve custom metadata fields attached to a file.
- Look up people and faces associated with a media file.
Metadata Editing
- Update a file's rating, flagged status, file name, model release status, and property release status.
- Update any IPTC/XMP metadata field on a media file (keywords, description, copyright, etc.).
Trash & Recovery
- List files currently in the trash.
- Restore trashed files to a destination gallery of your choice.
Shared Libraries
- List shared organizations accessible to your account.
- Browse and copy media from shared organization galleries into your own library.
Portal
- Fetch the top-level collections and galleries published to your Portal.
General
- Read what you are currently viewing in the library.
- Read which assets you currently have selected in the library grid.
- Answer questions about PhotoShelter features and capabilities using built-in documentation.
-
Search the web for real-time information, current events, or anything outside its knowledge.
What the Assistant Cannot Do
- Upload files — the assistant cannot upload new media to your library.
- Delete files or containers — it cannot move items to the trash or permanently delete them.
- Create smart (saved-search) galleries — it can read and rename them, but cannot create them or add files to them manually.
- Edit video — video trimming, clipping, and export are available in PhotoShelter for Brands, but the assistant cannot control those workflows yet.
- Content Derivatives — the assistant cannot create or manage channel-ready content derivatives.
- Interact with Socialie — the assistant is aware of Socialie but cannot take actions within it (e.g. sharing, suggesting content, or managing TikTok distribution).
- Modify Portal settings — it can read portal contents but cannot change access permissions, layouts, or other portal configuration.
- Manage users or account settings — it cannot invite users, change roles, or modify organization settings.
- Manage workflow automations (Formulas) — it cannot create, edit, or trigger automation rules.
- Provide pricing information — for any questions about plans, upgrades, downgrades, or billing, please contact support@photoshelter.com.
Tips for Best Results
- Be specific: "Find photos of the 2024 championship game tagged with 'basketball'" works better than "find sports photos."
- Refer to files or galleries naturally: phrases like "this gallery," "the selected files," or "the file I'm looking at" all work — the assistant will check your current view or selection automatically.
- For multi-step tasks (e.g. copying several galleries at once), the assistant will ask you to approve all actions upfront so you don't have to confirm each one individually.
The assistant is continuously improving. Capabilities may expand over time as new tools become available.