Microsoft Edge now automatically fills in missing image alt text for screen readers on Windows, macOS, and Linux, to be read aloud for blind or low vision users
Ideal for blind or low vision web users — More than half of the images on the web are missing alt text, according to Microsoft.
Context & Ripple Effects
This is Microsoft productizing a capability it has been building since 2020, when it detailed an Azure image captioning AI meant to improve accessibility across Word, Outlook, and other apps. Edge now applies that same idea at the browser layer, filling in alt text for the more than half of web images Microsoft says ship without any description.
The move sits alongside a broader accessibility stack: Immersive Reader's Read Aloud features opened to developers via Azure, and more recently Microsoft brought Be My Eyes' AI description tech into its Disability Answer Desk. Edge is becoming the delivery vehicle for that stack on the open web.
First-order effects
- Blind and low-vision Edge users on Windows, macOS, and Linux immediately get spoken descriptions for images that previously rendered as silence — addressing the majority of unlabeled images rather than waiting on publishers to fix them.
- Publishers who skipped alt text effectively outsource the job to Microsoft's model, with their images described by Edge's captions whether or not they wrote any.
Second-order effects
- Rival browsers face pressure to match the feature, since screen reader users can now pick a browser based on how well it handles unlabeled images — turning accessibility into a competitive differentiator rather than a compliance checkbox.
- If auto-generated captions prove good enough, some developers have less incentive to author alt text themselves, shifting quality control from site owners to browser vendors' models.
Third-order effects
- Assistive access to the web is drifting from depending on what publishers label to what platform AI infers, making browser vendors de facto arbiters of how much of the web is describable to non-sighted users.
- The pattern — captioning models moving from Office apps to the browser to customer service — points toward accessibility becoming an embedded AI layer across Microsoft's products rather than a per-app feature.
The trend: Browsers are absorbing accessibility work that publishers neglected, with AI-generated descriptions replacing author-supplied alt text as the default fallback.