Twitter Tests New Misinformation Labels - More Visible Warnings

Twitter's Enhanced Approach to Misinformation Labeling

Twitter is evaluating modifications to its system for providing context to tweets containing misleading information. These changes are being considered for content that, while inaccurate, doesn't meet the threshold for complete removal from the platform.

New Visual Cues for Misinformation

The company announced the testing phase via Twitter on Thursday, accompanied by a visual representation of the redesigned misinformation labels. This initial test will feature labels with color-coded backgrounds, significantly increasing their visibility within users’ feeds and enabling quicker comprehension through visual indicators. A subset of users will begin experiencing these changes this week.

Tweets identified by Twitter as “misleading” will be highlighted with a red background. This will be accompanied by a concise explanation and a restriction preventing users from replying to, liking, or sharing the post. Content deemed less actively misleading will receive yellow labels.

Increased Clarity and Access to Verified Information

In both instances, Twitter is emphasizing the ability for users to click on the labels to access verified information pertaining to the topic. Currently, this focuses on information related to the ongoing pandemic.

A Twitter spokesperson explained to TechCrunch that users participating in this limited test should anticipate a more impactful experience from the labels themselves. This impact will be achieved through the use of clear language, symbols, and colors designed to provide context not only about the label itself, but also the information being viewed.

Impact of Design on User Interaction

Twitter’s research indicates that even minor alterations in label design can influence user behavior. For example, a test utilizing a pink label variation resulted in increased clicks to authoritative information provided by Twitter. However, it also led to a rise in quote-tweets, potentially broadening the spread of the content.

The company reports that numerous variations of the label’s text, colors, and symbols were tested before arriving at the current design.

Responding to User Feedback

These adjustments are a direct response to extensive public feedback, which highlighted the need for misinformation labels to be more prominent amidst the constant flow of tweets. Similar criticism has been directed towards Facebook’s misinformation labels, which have been accused of being too subtle and ineffective in preventing the dissemination of potentially harmful information.

Evolution of Twitter’s Content Labeling System

Twitter initially implemented content labels to identify “manipulated media” – images and videos intentionally altered to deceive, such as the widely circulated doctored deepfake of Nancy Pelosi in 2019. In May of last year, the use of labels was expanded to address the surge of COVID-19 misinformation that emerged during the pandemic.

Furthermore, a month ago, the company introduced specific labels for vaccine misinformation and implemented a strike-based system within its policies.

A Proportional Response to Misinformation

Twitter aims to develop a comprehensive toolkit for responding to misinformation in a manner proportionate to the potential for real-world harm.

As a Twitter spokesperson stated, “not all misleading claims are equally harmful.” They illustrated this point by contrasting the danger of recommending bleach as a COVID-19 cure with the relatively benign impact of sharing a viral image of a shark in a flooded street falsely attributed to a hurricane.

Content Moderation Options

Labels represent just one of the content moderation strategies Twitter has developed over the past two years. Other methods include warnings requiring a click-through confirmation and pop-up messages intended to discourage impulsive sharing of inflammatory content.

When a decision is made not to remove content entirely, Twitter employs a range of enforcement options:

• Warnings
• Misinformation Labels
• Restrictions on engagement (likes, replies, shares)

In most cases, the company utilizes a combination of these approaches.

The spokesperson concluded, “we believe investing in a multi-pronged approach will allow us to be nimble and shift with the constantly changing dynamic of the public conversation,” acknowledging the complex challenges of addressing misinformation effectively.