Chrome’s New Gemini “Skills” Are Really About Locking In the AI Habit
Google’s latest Chrome feature, Gemini Skills, looks mundane on the surface: reusable prompts you can save and launch with one click. But this is less about convenience and more about training hundreds of millions of people to build a daily habit around Google’s AI. Once your workflow lives inside the browser sidebar, switching to a rival model—or even a rival browser—becomes a lot harder. In this piece we’ll look at what Skills actually are, how they fit into the broader AI assistant battle, what they mean for European users, and why this “small” feature is strategically very big.
The news in brief
According to Ars Technica, Google is rolling out a new feature in desktop Chrome called Skills, tightly integrated with the Gemini sidebar. Skills are essentially saved prompts: instead of typing the same instruction into Gemini each time, users can store it as a named action and trigger it with a click.
Skills are synced across devices via a Google account. While in the Gemini panel, typing a forward slash (/) or clicking a plus button opens a list of saved Skills, which can then run on the current tab or across multiple tabs. Google is also shipping a Skills Library with pre-made templates that users can add and customize.
The feature doesn’t change what Gemini can do technically; it makes those tasks faster to invoke. Security rules remain: actions like sending messages or adding calendar events still require explicit confirmation. The rollout starts in Chrome with US English language settings, and it’s available without a paid AI plan, as part of Gemini for Chrome.
Why this matters
On paper, Skills are just “bookmarks for prompts.” In practice, they are Google’s bridge from casual AI querying to embedded, repeatable workflows.
Most people don’t use chatbots like developers or power users. They remember one or two prompts, maybe ask for a summary or a quick rewrite, and move on. The friction of re‑typing or digging through notes for that clever prompt is exactly what stops AI from becoming a default tool. Skills directly attack that friction.
The immediate winners are:
- Productive power users who already lean on Gemini in Chrome: summarising dense PDFs, building comparison tables from many tabs, or turning messy text into structured output.
- Google, which gets users to encode their workflows into the Chrome + Gemini stack. Once you’ve curated 20 Skills that automate your browsing tasks, the cost of switching to Edge + Copilot or Safari + ChatGPT jumps dramatically.
The potential losers:
- Independent AI front‑ends and browser extensions that built a niche around “smart prompts” or workflow templates.
- Competing models that become harder to adopt because your muscle memory and saved Skills are bound to Gemini in Chrome.
The feature also nudges AI from “creative assistant” into automation tool. When a Skill becomes part of your daily routine—"clean up this CRM export", “digest this long legal page”, “extract numbers from financial reports”—you’re not experimenting anymore, you’re delegating.
That raises questions: How transparent is the automation? How easy is it to audit what a Skill is doing, or to move those workflows to another provider? On those fronts, Google is only taking baby steps.
The bigger picture
Skills sit at the intersection of three overlapping trends.
1. AI agents are getting more task‑oriented.
OpenAI’s GPT “custom versions”, Microsoft’s Copilot Studio flows, and countless “AI agent” startups all try to turn one‑off prompts into persistent tools. Google’s Skills are the same idea, but inside the browser, not in a separate app or studio. That matters because the browser is where knowledge work actually happens.
2. The browser is becoming the AI control center.
Microsoft pushed this first with Edge and Copilot. Arc, Opera, Brave and others followed with their own sidebars and summarisation tools. Chrome is late, but it has scale: more than 60% global market share means that any AI habit formed inside Chrome has outsized impact. Skills transform Gemini from “one more chatbot tab” into a sort of command palette for the web.
3. Prompt engineering is turning into product design.
Until now, good prompts lived in screenshots, Notion docs, or newsletters. Skills turn them into clickable, shareable affordances. That opens a door to a new ecosystem: libraries of specialised Skills for lawyers, students, analysts; marketplaces run by Google or third parties; maybe even enterprise‑controlled Skills deployed company‑wide.
We’ve been here before. Browser extensions, then cloud bookmarks, then password managers slowly made the browser the center of digital life. Skills are the AI era’s continuation of that arc: they store not just your credentials or links, but your ways of thinking through tasks.
Competitively, this increases pressure on:
- Microsoft, which must keep Copilot in Edge clearly more powerful to justify pushing users away from Chrome.
- OpenAI, which is strong on raw model quality but weak on tight OS‑ and browser‑level integration.
Whoever owns the default assistant in the default browser will have a structural advantage in the AI race—even if their model is not the absolute best.
The European / regional angle
For European users, Skills land in a regulatory and cultural environment that is very different from Google’s home market.
First, data protection. Under GDPR, any feature that potentially processes personal data—like a Skill that reads pages from your banking site or internal company portals—must be transparent about what’s sent to Google’s servers, where it’s processed, and for what purposes. Enterprises in Europe will demand clear controls: can Skills be restricted to on‑prem or EU‑hosted processing for sensitive workflows? Can admins disable certain Skills categories entirely?
Second, the Digital Markets Act (DMA) puts Chrome and Gemini under special scrutiny as “gatekeeper” services. Bundling powerful AI features into the dominant browser will attract regulators’ attention. Google will need to show that:
- Gemini Skills are optional and not required for a “normal” Chrome experience.
- Competing assistants can, in principle, get equivalent hooks into the browser, or at least are not unfairly blocked.
Third, Europeans are generally more privacy‑conscious. That may slow down adoption of heavily automated browsing workflows—especially in sectors like healthcare, law, and public administration. Local players (for example, browser vendors with EU‑based AI backends, or regional cloud providers offering compliant assistants) can position themselves as “AI, but on European terms”.
Finally, language support will be critical. The first rollout targets US English. For Skills to matter in the EU, Gemini needs robust capabilities in German, French, Spanish, Italian, Polish and smaller languages, plus Skills libraries that reflect local services, regulation and culture—not just US‑centric templates.
Looking ahead
Skills are almost certainly version 1 of a much larger idea.
Today, a Skill is basically a saved text prompt. Expect that to evolve into richer, semi‑programmable workflows:
- Conditional logic ("if the page is a PDF, do X; otherwise do Y").
- Chained actions ("summarise these three tabs, then draft an email in Gmail").
- Shared organisational Skill sets managed by IT teams, with policy controls.
We should also watch for:
- Integration with Android and ChromeOS: one Skill to process a page on desktop and continue on mobile.
- Marketplaces: if Google opens a verified Skills gallery, we’ll see a repeat of the Chrome Web Store era, with all its innovation and security headaches.
- Regulatory responses in the EU: the more “agentic” Skills become—taking actions on your behalf—the more they’ll look like something regulators want to certify or at least closely supervise.
Timeline‑wise, expect Skills to spend 2026 as a semi‑hidden power feature for enthusiasts and early adopters. The real inflection point will come when:
- Language and regional support broadens.
- Google starts surfacing Skills proactively—suggesting them contextually rather than waiting for users to click.
At that point, Chrome ceases to be “just a browser” and becomes the default UI for Google’s AI agent. That’s when both competitors and regulators will have to react in earnest.
The bottom line
Skills make Gemini in Chrome feel less like a chatbot and more like a toolbox. That’s good for productivity, but it also deepens dependence on Google’s ecosystem and raises fresh questions about transparency, portability and competition—especially in Europe. If your daily work increasingly flows through a set of AI‑driven Skills, who really controls your workflow: you, or the platform? Before we all standardise on Chrome as our AI cockpit, that’s the question worth asking.
