Dev tools

Microsoft Edge automates extension review checks after AI-assisted submissions overwhelmed its pipeline

September 9, 2026 at 10:30 AM PT

Microsoft Edge browser hero image

Image: Microsoft

Why it mattersAn extension developer who used to wait weeks for an Edge Add-ons review now has a faster and more predictable review pipeline, and Featured recognition catches up with a fixed bug in half a month rather than whenever a human reviewer next runs the audit.

The Microsoft Edge team said on 2026-09-08 that AI-assisted coding has raised the volume of extension submissions past what its review pipeline can absorb, and it has responded by automating the parts of the review that repeat. The post is under the byline of the Edge Team on the Microsoft Edge Blog.

The team says rapid adoption of AI-assisted coding is letting developers build extensions faster than before, and that more developers are building, iterating, and submitting them to the Edge Add-ons site. An expedited review process introduced last year for high-quality and high-value extensions has since been overwhelmed by the same growth: submission volumes kept climbing, and turnaround times went up.

Where the automation lands

The change is that many of the repeatable validation checks in the review process now run without a person, and the review pipeline itself has been streamlined. The team explicitly says the automation does not change review standards and does not remove any of the checks an extension must pass. What it does is identify known policy violations and security issues more consistently, and free reviewers to spend their time on the complex cases that need human judgment.

The Edge post does not say whether the automation itself uses AI. It also does not give a before-and-after number on review times, only that the result should let bug fixes and new features reach users sooner. For a developer whose last submission sat in the queue, the direct promise is a shorter wait, with the same review bar.

The context matters here. Microsoft's Exchange team said in mid-August that AI had found so many bugs that the team could not deliver a Cumulative Update on time, as reported by The Register. The Edge team is now the second Microsoft group publicly citing an AI-driven surge as the reason a shipping process is under strain. In both cases the response is more automation rather than more people.

Featured badge refreshes go to a fixed cadence

The second change is on how Edge recognises quality on the store. The Featured badge marks extensions that meet the team's Best practices guidance, and the evaluation now runs against more than 60 quality signals across multiple dimensions of an extension. The team says much of that evaluation is already automated so the badging is more consistent and scales with the store.

Featured badges will now refresh every 15 days. That is a fixed cadence for a check that had been re-run less predictably, so a developer who fixes an issue or ships an improvement can see the badge follow within two weeks rather than at whatever interval the audit next happened to run.

The user-facing side of the same change is that the Featured list on the Edge Add-ons site reflects the current quality of the extension rather than the state it was in at the last badge sweep. An extension that has slipped will lose the badge on the next refresh, and one that has improved will gain it on the same schedule.

Source

Source: Microsoft Edge Team

This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.

More from AI News

GitHub adds a ruleset that blocks pull requests from merging while a secret they introduce is still open

GitHub said on 2026-09-09 that a new repository rule blocks a pull request from merging if any secret scanning alert introduced by its commits is still open, closing a gap that push protection did not cover for changes coming in through a fork or a rewritten branch.

Source: Vendor blogDev tools

Ponytail makes a coding agent check seven things before it writes any new code

Ponytail is an MIT-licensed rule set that makes a coding agent work through a seven-step ladder of cheaper options before it writes new code, and it has reached 127,785 stars on GitHub.

Source: Product HuntDev tools

CodeRabbit measured GPT-6 Astra at 61.3% bug coverage in code review, at 2.5 times the cost of Sol

CodeRabbit published an early evaluation of GPT-6 Astra on code review, reporting 61.3% actionable bug coverage against 59.0% for GPT-5.6 Sol, with a much wider gap on cross-file reviews and a token price 2.5 times higher.

Source: Hacker NewsDev tools