A new resource or landing page can leave dozens of older videos pointing people to yesterday’s offer. Updating a video archive is a good AI task because the goal is repetitive while the surrounding content varies. Each description still deserves a careful, specific edit.
Write the edit rule before touching the archive
Imagine a tutorial channel replacing a discontinued checklist with a new starter guide. Some videos mention the old checklist in the opening paragraph. Others link to it under a resources heading. A few never mention it. A global text replacement could miss links or insert an irrelevant promotion.
Give the assistant the old destination, the new destination, the approved call to action, and the selection criteria for eligible videos. State what must survive: chapter timestamps, sponsor disclosures, affiliate notices, credits, and unrelated links. Ask it to flag videos where the new offer does not fit the subject.
Create a review sheet and a restore copy
Before editing, export the exact current descriptions with their video IDs and URLs. Keep that original file untouched. A separate review sheet should show the proposed description, a short change explanation, and whether the entry is ready or needs attention.
A connector or authenticated editing method is required to change live YouTube content. Writing a prompt does not provide that access. If you do not have a supported integration, AI can still prepare the revisions for you to apply in YouTube Studio.
- Define the eligible videos and save their original descriptions.
- Generate proposed revisions while preserving unrelated text exactly.
- Review a small sample covering the different description formats.
- Authorize the specific batch to update, then verify the saved descriptions against the proposals.
Make tracking readable
If you use campaign parameters, pick a consistent naming convention before making the links. For example, a campaign name can identify the guide launch while a content value identifies the individual video. Document the convention so the traffic report is understandable later.
Open sample destinations and check that the page loads with its parameters intact. If you use a short-link service, verify both the short URL and the final destination. Never generate hundreds of links before checking that the first few point to the right resource.
Treat the live update as a separate stage
A review sheet is not proof that YouTube saved the changes. Ask for a completion log listing successfully updated IDs, skipped videos, and failures. For representative videos, compare the saved live description with the approved revision and confirm that the chapter markers and disclosures remain intact.
Make the operation safe to repeat. If the new call to action already exists, the next run should recognize it instead of inserting it again. Keep the original export until you are satisfied with the results and know how you would restore an individual description.
The same method applies to a resource link that changes every year or an archive of event recordings. Separate selection, revision, and publication so you can inspect the result at the point where a mistake would be cheapest to fix.
A prompt to make it your own
Prepare description updates for these YouTube videos: [IDs or selection criteria]. Replace references to [old resource] with this approved wording and destination: [copy and URL]. First export the complete original descriptions with video IDs and URLs. Preserve chapter timestamps, disclosures, credits, and unrelated text exactly. Flag videos where the replacement is not relevant. Use this tracking convention if applicable: [rules]. Return original and proposed versions with a change explanation for each video. Check for duplicate calls to action and test sample destinations. Do not publish yet; I will review the proposed batch first. After an explicitly authorized update, report saved, skipped, and failed IDs.
A bulk edit should be a collection of traceable, reversible changes, with a clear record of what actually saved.
Use-case inspiration: 9x’s ChatGPT Work video and companion article. Examples, instructions, and prompts written for AI Prompts.
Use the prompts with the tools and permissions available in your AI environment. Some workflows require connected accounts or uploaded files.