What I explored
How to give users temporary control over repeated exposure to heavy topics without turning the product into a censorship or avoidance tool.
Chrome Extension / Product Design
A Chrome Extension that helps users soften unwanted news topics without fully disconnecting from the news.
5 min read
05/2026
How to give users temporary control over repeated exposure to heavy topics without turning the product into a censorship or avoidance tool.
Filtering is a trust problem as much as a technical one. Users need to know what was softened, why it happened, and how to reveal it.
I would test the smallest meaningful blur target earlier, because filtering one keyword can accidentally hide much more content than intended.
News browsing can become emotionally repetitive. A user may want to stay informed, but not repeatedly process the same heavy topics across headlines, sidebars, most-read lists, and recommendations.
The problem was not that news sites show difficult content. The problem was the lack of moment-to-moment control. Sometimes the user does not need a permanent block. They need a softer boundary: not right now.
Product principle: FilterNews should reduce repeated exposure without hiding the fact that something was filtered. Relief should not come at the cost of control or awareness.
The design sits close to cognitive load and emotional regulation. When the same kind of stressful topic appears repeatedly, the cost is not only reading one article. It is scanning, interpreting, reacting, and deciding whether to keep going.
FilterNews gives the user perceived control over that exposure. The important part is practical: the user can soften a topic, see the reason, reveal the content, and decide whether the filter should apply only on specific sites or only for a limited time.
FilterNews uses soft filtering instead of deletion. Matching content is blurred or dimmed, a small label explains why it was filtered, and the user can reveal it when they choose.
This keeps the experience honest. The product is not pretending the news disappeared. It is helping the user decide whether now is the right moment to engage with it.
Feature hierarchy: the timer supports the main product promise: temporary control. Cat Mode is intentionally lighter. It adds personality, but it should not carry the value proposition.
The hardest part was not detecting words. The harder design problem was deciding what should happen when the filter is wrong, too broad, or surprisingly effective.
A keyword can misunderstand context. In one test, the word "accident" appeared in a banner telling readers to send photos or news tips and to call 112 first in an emergency. The filter detected the word and blurred the banner, even though it was not an accident article.
That made the trust issue obvious: the user needs to know why something was filtered, and they need an easy way to reveal it. Without transparency, the product starts to feel arbitrary.
If a filtered word appears inside a large page section, the extension can accidentally blur the whole section even when only one headline matched. That makes the product feel heavier than the content it is trying to soften.
The design implication is simple: the filter should target the smallest meaningful unit. On a news site, that usually means an article card, headline row, sidebar item, or recommendation block.
With only a few custom words active, a surprisingly large part of the news homepage became filtered. That was the strongest product signal in the project.
Emotional overload is often caused by repeated exposure, not one single article. The same topic can appear in the main story, a sidebar, a most-read list, and several recommendation blocks. FilterNews is useful because it works across that repetition.
Learning: a small number of filters can make a large difference when the page repeats the same topics in many places.
Cat Mode is a deliberately small opt-in layer. If the user turns it on, filtered areas can be replaced with calm cat images instead of blurred content.
I would frame it as a playful escape hatch, not a serious coping feature. Most users probably would not keep it on every day, and that is fine. Its job is to make the product feel less clinical while keeping the default experience quiet and useful.
This project taught me that filtering is not only a technical problem. It is also a trust problem.
If the filter hides too much, it feels restrictive. If it hides too little, it does not help. The core design principle is to keep the user in control: explain the reason, keep the content recoverable, and let filtering be temporary or site-specific.
The product direction I would keep is the softer framing. FilterNews should not promise to remove reality from the internet. It should help people stay informed without being forced to repeatedly process topics they do not have capacity for right now.