Another Microsoft team admits it’s struggling to handle flood of AI-generated code

Microsoft's Edge team has admitted that it is struggling to keep up with the influx of AI-generated code. The team revealed that while AI-assisted coding is enabling developers to build extensions faster than ever, it has also led to more extensions being built, iterated, and submitted. Microsoft must now vet these extensions before allowing users to install them, which has increased the workload for the Edge team.

Initially, the company tried to address this issue by introducing an expedited review process for high-quality and high-value extensions. However, the submission volumes continue to rise, and the review pipeline is facing additional strain, causing an increase in extension review turnaround time. To tackle this problem, the Edge team has introduced automation for many of the repeatable validation checks in the review process and streamlined how reviews move through the pipeline.

This automation does not compromise the review standards or the checks an extension must pass; instead, it helps identify known policy violations and security issues more consistently, allowing reviewers to focus on complex cases that require human judgment. The team believes this change will enable it to approve and update extensions more quickly, refresh the "Featured" badge applied to high-quality extensions every 15 days, and provide faster feedback to developers.

Consequently, users will benefit from a more up-to-date view of extensions that meet quality standards.