Gems Are Retiring, Skills Take Over: Turn Your Company's AI Helpers into Portable AI Skills
If your team has ever built a "Gem" in Gemini, such as a helper that writes product copy or an assistant that formats meeting notes the company way, it is time to plan a move. Google has announced that Gems will be retired and replaced by "skills": Gems disappear from personal accounts in November 2026 and from Workspace business accounts in March 2027. This is more than a rename. It reflects a bigger trend: AI skills are becoming a portable, cross-platform format for work instructions. Here is what is changing, and how companies can use the moment to turn their know-how into AI skills they can genuinely reuse.
Google Is Retiring Gems: Timeline and Differences
According to Google's official help page, the retirement dates depend on account type:
- Personal accounts: removed in November 2026, along with Gems by Google Labs and the mini-app tool Opal.
- Workspace business, enterprise and non-profit accounts: removed in March 2027.
- Workspace education accounts: removed in June 2027.
Existing Gems will be converted to skills automatically when they are removed, or you can convert them early by hand. The biggest difference is how they are used. A Gem is a dedicated chat room you have to enter first. A skill can be called up in any conversation with a command, applied automatically when Gemini judges it relevant, and combined with other skills.
There are limits for now: skills do not yet support Canvas or Deep Research, sharing and Drive files are still on the way, and work and school accounts will get skills later.
Why AI Skills: From a Dedicated Chat to an On-Demand Playbook
Behind this change is a shared format taking shape. In December 2025, Anthropic published "Agent Skills" as an open standard: a skill is a folder containing a SKILL.md file with a name, a purpose and step-by-step instructions, optionally bundled with templates, reference material or small scripts. The AI reads only the name and description by default and loads the full content when it is needed, so many skills can stay available without slowing things down.
Google's skills can also import SKILL.md files. For businesses, the practical meaning is clear: the AI skills you write are less likely to be locked into a single platform. Use them in Gemini today, move to another AI tool tomorrow, and the core instructions can come with you.
Turning Your Company's AI Helpers into AI Skills: Four Steps
Rather than waiting for automatic conversion, use this moment to tidy up the AI helpers scattered across your team.
1. Take inventory of Gems and prompts
List the Gems your team uses, the long prompts saved in notes, and the "ask AI this way" tips passed around by word of mouth. Many companies discover that three people have written three versions of the same thing.
2. Pick the work worth turning into skills
The best candidates are tasks that are repetitive, have clear standards and are costly to get wrong: rewriting copy in the brand voice, turning client requirements into a quote format, checking a press release against style rules.
3. Write it as if onboarding a new hire
A good skill reads like a handbook for a new colleague: the purpose, the steps, the criteria for judgment, and examples of good and bad output. We covered the same principle in brand guidelines AI can read: the more specific you are, the less the AI improvises.
4. Assign an owner and a review date
Every skill needs an owner and a regular check for outdated content. When brand rules or product lines change, the skill has to change too.
What Makes AI Skills Good Is Still People
Skills make AI more reliable, but writing them takes judgment. Which experience is worth writing down, which exceptions need a warning, and when the AI should stop and hand over to a person: these decisions belong to people who know the business.
Keep two ground rules in mind. Never put passwords or customer personal data in a skill file. And for decisions involving external commitments, pricing or contracts, a skill can help prepare the work, but a person makes the final call.
Conclusion: Turn Know-How into an Asset
Retiring Gems looks like a minor product change, but it is a useful reminder: if your company's way of working lives only in someone's head or in one platform's chat room, it can vanish with a staff change or a tool switch. Turning it into clear, portable AI skills is the first step toward making AI a real member of the team. That is the division of labor TitApeX believes in: AI executes flawlessly, people define what "good" means. If you want to turn your processes and brand standards into skills AI can use, see our services or get in touch.
Comments