A report commissioned by the Browser Choice Alliance (BCA) and compiled by the behavioral science organization Behavioural Insights Team (BIT) outlines 10 types of dark patterns that can get in the way when users try to choose a browser other than Microsoft Edge on Windows. The 46-page report, published on the BCA website in October 2026, looks beyond browser installation to everyday use and OS updates. The core problem: even after a user picks a different browser, that choice does not necessarily carry over to every Windows feature.
A default browser is the standard browser used to open ordinary web links. However, the Windows-wide default setting is managed separately from the setting that determines where links open inside Outlook and Teams. If those apps open a link in Edge, that does not mean the Windows default browser itself has been changed. The app used to open files such as PDFs also depends on separate file associations.
The report is a commissioned analysis that checks materials supplied by the BCA against existing behavioral science research. It is not a new survey based on random sampling of Windows devices around the world. Reading it alongside Microsoft's official explanations shows that the issue should not be framed simply as "being switched back to Edge." What matters is distinguishing which settings and features reflect a user's choice, and where that choice does not reach.
Ten tactics, from browser installation to OS updates
The BIT report treats interface designs that interfere with users' intended choices as dark patterns. At the installation stage, it criticizes prominent banners that appear when users search for competing browsers, as well as mechanisms in which changing the default browser does not fully switch file associations. S mode in Windows is also part of the analysis, but BIT notes, based on Microsoft's submission, that it was discontinued on new devices manufactured after September 30, 2025. The situation for existing devices and new devices therefore needs to be considered separately.
On everyday use, the report examines how links behave when opened from Windows Search results, Widgets, and the lock screen. It also analyzes links opened from the "News and interests" feature in Windows 10, Outlook, and Teams as separate routes of steering. It further covers workarounds for avoiding Edge being blocked, and a mechanism that pulls tabs from other browsers into Edge at sign-in. The 10 tactics the report identifies are not ads or warnings on a single screen; they span multiple Windows features and Microsoft products.
On OS updates, the report criticizes designs in which the approval button on screens such as "Use recommended browser settings" is made prominent, while declining requires extra steps. However, the report does not claim that every OS update will necessarily revert the default browser to Edge without user consent. Nor does it imply that all 10 tactics apply simultaneously to every region or user.
Changing the Windows default browser doesn't switch every link
Microsoft itself explains how links from Outlook and Teams are handled. For eligible Windows 10 and 11 users, there is a feature that displays the linked web page side by side with the original email or chat in Edge, which Microsoft says is intended to reduce switching between apps. According to the company's official FAQ, this feature does not change the Windows default browser; it changes which browser is used when links are opened from Outlook or Teams.
The feature can be turned off in the settings of either Outlook or Teams, and the change applies to both apps. The location of the setting varies by app version, however. In the new Outlook, for example, it is changed under the file and link settings in General. Note that this is Microsoft's published description of a product feature, not a rebuttal issued in response to the BIT report.
To understand these differences, it helps to separate Windows file associations, link settings in Microsoft apps, and changes made for the European Economic Area (EEA). The table below compares the BIT report with two official Microsoft documents to show what each setting covers. It is not the result of new hands-on testing.
| Setting / route | What the change applies to | What needs separate checking |
|---|---|---|
| Windows-wide default setting | Links and file associations supported by the browser | Coverage varies by region and by the browser's own support |
| Individual associations such as PDF | The app that opens the specified file type | Not necessarily linked to a browser-wide change |
| In-app settings for Outlook and Teams | Where links opened from those apps are displayed | Independent of the Windows default; changes apply to both apps |
| EEA Search, Widgets, etc. | Links opened from Windows features Microsoft changed | Windows 10 "News and interests" is excluded |
In other words, changing the Windows default browser alone does not mean links from email and chat will also switch to the same browser. To confirm that a setting has taken effect, you need to distinguish not only the type of link or file but also which app or feature it was opened from.
In the EEA, the one-step default browser change has been expanded
Regional differences can be seen in Microsoft's official explanation of its Digital Markets Act (DMA) compliance. Published on June 2, 2025, and updated on July 25 of the same year, the document explains that in the EEA, when a browser that has completed the required registration is set as the default, more links and file associations can be switched in one step. The change expands the range covered beyond file types that were already included, such as htm and html.
Specifically, the ftp, http, https, and read protocols are covered, along with the htm, html, mht, and mhtml file formats. shtml, svg, xht, xhtml, and xml are also included, and PDF can be set separately. It is therefore inaccurate to describe the targets of the previous one-step change as "only http and https."
The same document says that links opened from the Bing app in Windows Search, Widgets, and the lock screen's Start Experiences will be changed to use the default browser. On the other hand, it says links from Windows 10's "News and interests" will continue to open in Edge. In other words, even within the EEA, not all Windows features have been unified under the same rules. Nor can it be said that these changes have been applied in Japan, and an explanation published in 2025 alone cannot establish how every device behaves in 2026.
How well is the burden of re-choosing actually demonstrated?
What emerges from these materials is that what matters is not only whether a button for changing browsers exists, but how much effort it takes for users to keep the state they chose. That said, this is an analysis in this article that draws on how the settings work and regional differences, not a conclusion from newly measuring the effect on actual users. The BCA, which commissioned the report, counts Google Chrome, Opera, and Vivaldi among its members. Waterfox is also listed through Wavebox, Midori, and BrowserWorks.
BIT's analysis relied on materials such as screenshots provided by BCA members in June 2026 and responses to regulatory inquiries. The report is explicit about the source and scope of these materials, stating:
"The findings in this document are based solely on materials provided by the BCA" (Japanese translation)
This limitation indicates the source of the materials used to analyze Windows screens and behavior. As for the effect on user behavior, BIT draws on existing behavioral science research. It also discloses that it used AI as an aid in searching and organizing literature and in drafting, and says the output was reviewed by researchers and senior staff. The user personas in the report are fictional figures used to illustrate the effects, not actual experiment participants. The report should therefore be seen as a combination of analysis of supplied materials and a synthesis of existing literature, not a peer-reviewed experimental study.
In its review of existing research, BIT assesses the evidence for time and cognitive burdens caused by dark patterns as relatively strong. Evidence for the effect on user autonomy is rated strong to moderate, and for the effect on the ability to make informed choices, moderate to limited. Evidence on frustration and loss of trust is more limited still. The report does not offer a quantitative estimate of how many users actually changed their browser choice because of individual tactics on Windows.
The report also says the effects on older people, people unfamiliar with digital devices, and people short on time are not uniform. BIT itself acknowledges that preset options and warning screens can be useful in smoothing setup or avoiding real risks. What it objects to is cases where such mechanisms obstruct users' intended choices without legitimate justification, or where multiple steering tactics overlap and increase the burden of choosing.
Going forward, evaluating how freely users can choose a browser on Windows will require checking whether settings take effect not only for ordinary web links but also for each route, including PDF files, links in email and chat, and OS search results. It will also be important to hold conditions such as region, Windows version, and app type constant, and to keep verifying whether the chosen state persists after OS updates. Clarifying where settings can be changed and how far those changes actually reach would give a more accurate picture of how many times users have to re-select their browser.