WordPress 7.0 'Armstrong': The Release That Closes a Decade-Long Gap

Multiple cursors in the same document, in WordPress core. That is the headline of 7.0 ‘Armstrong,’ which shipped May 20 with real-time collaborative editing, the milestone the Gutenberg roadmap has pointed at for a decade.

It is not the only big change in the box. The release also ships a Web Client AI API, one standard way for generative tools to plug into the editor. And it delivers the first comprehensive admin redesign since 2013. One version number, three real transitions.

One note before the practical guidance. This release was supposed to ship in April. It slipped a month because the collaboration architecture needed rework, and the project said so plainly and shipped when it was ready. Credit where it is due.

Who collaboration actually helps

Real-time co-editing builds on the block-level commenting that arrived in 6.9. Together they finally give WordPress the review-and-write-together workflow that pushed so many content teams into Google Docs.

Sort your adoption by team shape. Newsroom-style teams, campaign groups, and any content operation where drafts pass through many hands will feel this immediately. The draft-in-Docs-then-paste pipeline finally has a retirement date.

Solo-editor sites can shrug. The feature costs them nothing and changes nothing. The middle tier should let actual workflow pain set the pace, not feature excitement.

The usual first-version caveats apply. Editor-adjacent plugins need compatibility checks, very large documents deserve testing, and the point releases will sand the edges. Pilot with one team before rolling it wide.

The AI API is really a governance feature

The Web Client AI API sounds like a developer detail. It is actually the answer to a governance problem we have flagged for two years.

Until now, every AI plugin brought its own provider deal and data flows. Each also brought its own unwatched prompt box. The new API replaces that sprawl with one standardized, auditable surface through which generative features reach the editor.

That means the configuration work is yours to do, deliberately. Choose your providers. Map permissions to roles. Turn on logging. If your organization wrote AI content policies over the past few years, this is the layer that finally enforces them. Do the setup before the plugins choose defaults for you.

Budget for the retraining, seriously

The admin redesign is the change every user will feel first, and its real cost is the one nobody budgets.

A comprehensive redesign of the most-used screens in your institution retrains every author’s muscle memory. Plan a pilot group, updated documentation, and short training sessions timed to your rollout. Expect a help-desk wave, and schedule the rollout for a quiet season rather than your busiest one.

Also worth remembering: every internal guide your team has written since 2013 contains screenshots that are now wrong. Put the documentation refresh on the same calendar.

How to run the collaboration pilot

For teams ready to try the new editing, here is the rollout we are using.

Pick one editorial team with a real multi-author workflow, not a volunteer group of enthusiasts. Move one recurring content type into WordPress drafting for a month: the newsletter, the news posts, whatever passes through the most hands. Keep the old Docs pipeline available as the escape hatch, but ask the team to start in WordPress.

Watch two things. Whether the commenting and co-editing actually replace the Docs conversation, and where people fall back to old habits. The fallbacks tell you what to fix before the wider rollout, and they are usually about notifications and mobile editing rather than the core features.

Then set a retirement date for the paste pipeline and announce it. Workflow habits outlast feature gaps by quarters, and a dated cutover is the only thing that actually ends them. Our own team is on week three of exactly this pilot, and the early verdict is that the feature is good and the habit change is the real project. Budget for the second, not the first.

A good month for a platform people wrote off

Two years ago, the governance turmoil around WordPress had plenty of people predicting decline. This release is a strong argument against that reading.

A solid, well-built 7.0 shipped close to schedule, through the turbulence. It closes the collaboration gap that fueled a decade of ‘WordPress versus modern tools’ hedging. The concerns about project governance have not disappeared, and we still track them. But delivery is the best evidence there is, and this is real delivery.

Platforms are resilient in proportion to their communities, not their org charts. The proof shipped on May 20. Pilot the collaboration. Configure the AI layer on purpose. Budget the retraining. Then enjoy a release that was worth the extra month.