When 'No AI' Is the Feature: What LibreOffice Tells Us

LibreOffice just broke its own download records. The headline reason: it announced it has no AI features.

That's worth sitting with for a minute before you greenlight your next AI integration.

What Actually Happened

LibreOffice is a free, open-source office suite that competes with Microsoft 365 and Google Docs. Both of those products have been aggressively shipping AI writing assistants, autocomplete, and summarization tools. LibreOffice went the other direction and said, plainly, that it has none of that.

Downloads spiked.

This wasn't a fluke. A meaningful chunk of users actively chose software because it promised to stay out of their way. That's a real market signal, not a nostalgia blip.

The Gap Between AI Enthusiasm and AI Trust

If you follow AI industry coverage, it looks like everyone wants more AI everywhere. The surveys say adoption is rising. The venture dollars say the same.

But there's a gap between enthusiasm for the idea and trust in a specific product to do it well. Most users have now seen AI autocomplete confidently produce wrong information, AI assistants rewrite things they didn't ask to have rewritten, and AI features that activate when they didn't intend to trigger them.

Fatigue is real. People are making deliberate choices about where they let AI touch their work.

This matters for anyone building or evaluating AI agents for their business.

The Right Lesson (and the Wrong One)

The wrong lesson is "don't build AI agents." That's not what this story says.

The right lesson is that the way you integrate AI is a product decision with trust consequences. Features that act on data without clear user intent, that surface at the wrong moment, or that fail visibly tend to poison trust for the whole surface area. LibreOffice users aren't rejecting automation. They're rejecting automation that felt imposed on them.

AI agents built for business operations work differently. A well-scoped agent handles a specific, bounded job: processing inbound leads, routing support tickets, drafting first-pass reports from structured data. It runs in the background, on a trigger the operator controls, and its output is reviewable before it affects anything downstream. That's a different proposition from a writing assistant that rewrites your sentence mid-keystroke.

The failure mode to avoid is the same in both cases: the system acts in ways the user didn't anticipate and can't easily reverse.

What This Means for Evaluating Automation in Your Business

Before adding an AI agent to any workflow, it's worth asking three concrete questions:

Who controls the trigger? Does the agent run when a human decides it should, or does it fire automatically on conditions that might be wrong?
What's the blast radius of an error? An agent that drafts a Slack message for human review has a small blast radius. An agent that sends customer-facing emails without review has a large one. Scope accordingly.
Can the output be inspected before it matters? Reviewability isn't just about catching mistakes. It's about building the internal trust that gets adoption past the first week.

If you're not sure where an agent would actually save your team time versus where it would create anxiety, our free AI Opportunity Audit maps your highest-impact candidates from just your website. It's a useful starting point before committing to a build.

The Operators Who Will Win This Cycle

The LibreOffice story is a reminder that AI adoption isn't one-directional. Users push back. Trust is earned incrementally. The businesses that get durable value from AI agents are the ones treating integration as a trust-building exercise, not a feature-shipping exercise.

That means starting with automation that's clearly scoped, clearly useful, and clearly under human control. It means shipping something small that works well rather than something ambitious that surprises people. And it means being willing to keep some things AI-free if that's what earns confidence in the parts that aren't.

The market is telling you something. The question is whether you're reading it as a signal about your specific build, or just assuming it applies to the other guy.

Build Agents That Earn Trust

If you want to build automation your team will actually use, not another tool they route around, book a call and we can talk through what that looks like for your operation.