Apple’s grey list: Words and phrases your iPhone won’t help you spell
First reported by The Guardian ·
Your iPhone intentionally limits the words and phrases it will help you type, impacting how easily you can communicate certain ideas.
Apple employs a "grey list" of over 100,000 words and phrases that are intentionally excluded from its iPhone's predictive text and swipe-typing features. This list, discovered by independent researcher Derik Kauffman, prevents the device from suggesting or auto-completing terms related to drugs, sexual acts, and potentially sensitive political or social topics, such as "male chauvinist pig," "Trayvon Martin," and "Kashmir dispute." While Apple states this is to avoid inappropriate suggestions, Kauffman described it as a "light-touch form of censorship." The "grey list" affects the usability of swipe-typing, sometimes leading to unintended word changes for restricted phrases, like "United Ireland" becoming "United Iceland." This hidden list offers insight into the mechanisms behind Apple's autocorrect technology, which recently evolved to a more advanced, transformer-based system.
The existence of Apple's "grey list" reveals a deliberate, large-scale filtering of user input that goes beyond simple spell-checking. This indicates a proactive stance by Apple to control the narrative and user expression on its platform, particularly concerning sensitive or controversial topics. For users, this means that certain politically or socially charged phrases will require manual typing and may even be subtly altered by the swipe-typing function, potentially hindering direct communication or discussion.
This practice raises questions about algorithmic bias and content moderation within consumer technology, particularly as AI-driven text prediction becomes more sophisticated. Companies are making implicit editorial decisions through these lists, affecting what is easy and what is difficult to articulate. The discovery highlights the opacity of these systems and suggests that users may not be fully aware of the invisible barriers shaping their digital communication, prompting a closer examination of user control and platform responsibility.
AI-written summary. May contain errors.