The Guardian is reporting that Microsoft has installed NewsGuard by default in new installs of its Microsoft Edge mobile browser. The Daily Mail is one of the affected publications:

Visitors to Mail Online who use Microsoft Edge can now see a statement asserting that “this website generally fails to maintain basic standards of accuracy and accountability” and “has been forced to pay damages in numerous high-profile cases”.

Now, don't get me wrong: the Daily Mail is a horrendous excuse for a newspaper that, indeed, appears to bend the truth in order to further a toxic, conservative agenda. But let's zoom out a bit and examine this feature in the abstract.

NewsGuard assesses news websites manually, and gives them a rating for creditability and transparency. Based on that rating, it then assigns the website a red or green rating, indicating whether you should trust it or not:

These ratings are then displayed next to website content, and embedded into search engine results, via NewsGuard's browser plugin.

As a stand-alone plugin, this is probably fine. If you've made an active decision to install it, and you trust its editorial team to provide ratings, then great: you're informed about the process, you know that NewsGuard provides subjective ratings, and you've made the decision to overlay them over your web browsing experience.

As a default feature of a web browser, it's quite another story. For these users, it's a core part of their web experience. As far as they're concerned, this is a built-in feature, endorsed by Microsoft, that provides objective ratings on the content you browse. NewsGuard has been handed the ability to decide which content and information these users should trust.

These aren't just some do-gooder journalists. NewsGuard's advisory board contains the former head of the CIA and the first Secretary for Homeland Security. In light of this, some hypothetical questions to ask include: how might an independent website publishing the Pentagon Papers have been rated? What if a publisher is considered to be politically subversive while maintaining accuracy? In the wrong hands, could this mechanism suppress whistleblowers?

A web browser has no business telling you whether to trust the content you're accessing, except on technical grounds. If the web is to remain an impartial platform that supports freedom of speech, it cannot make value judgments on that speech. At least not by default: the extensions you install in your browser are up to you.

The whole point of the web is that it's decentralized, and anyone can become a publisher at any time. Yes, disinformation is a real problem. But it shouldn't be used as an excuse to put trust in the whole platform under central control. To do so introduces a real risk for the health of the internet, and, because freedom of speech is a prerequisite for it to function, for the health of democracy.