# iPhone Predictive Keyboard Blocks Controversial Phrases Researcher identifies over 100,000 restricted phrases in Apple's iOS predictive text system. - **Source:** [ANSA – Top News](https://www.ansa.it) - **Published:** 2026-11-09 - **Topic:** [technology](https://brifing.si/c/technology) - **Original article:** - **On Brifing:** https://brifing.si/s/12d4136e-7a19-4c2d-94c8-7a18b0b0e7b6 ## Restricted Predictive Text Discovery Researcher Derik Kauffman identified over 100,000 phrases restricted by the iPhone predictive keyboard. These terms are blocked from automatic completion and swipe typing features across multiple languages. While users can still manually type these words letter by letter, the iOS system refuses to offer them as suggestions, effectively limiting their visibility. ## AI-Driven Data Analysis Kauffman utilized advanced artificial intelligence tools, specifically OpenAI Codex and Anthropic Claude Code, to analyze the iOS system. By decoding internal tables stored within the operating system, he successfully converted hidden data into readable text. A second security analyst has since confirmed the existence and composition of this extensive restricted database. ## Scope of Blocked Content The list includes a wide range of politically charged and socially sensitive topics. Examples include phrases like 'Trump is a dictator', 'Covid is a hoax', and 'Crimea is Russian'. The database also flags controversial terms such as 'illegal immigrant' or 'left-wing conspiracy', alongside seemingly benign statements like 'Santa Claus does not exist'. ## Apple's Official Stance Apple responded to inquiries from Reuters by stating that the keyboard is designed to avoid generating potentially inappropriate or offensive suggestions. The company emphasized that the system is a safety feature rather than a total ban. Users retain the full authority to ignore or manually override these predictive suggestions during their daily typing. ## Technical Implementation Details The restriction mechanism functions by excluding specific strings from the predictive text engine's dictionary. When a user begins typing a blacklisted phrase, the iOS software may automatically modify the input or simply withhold the suggestion. This ensures that the device does not actively promote or facilitate the use of flagged terminology. ## Implications for User Autonomy This discovery highlights the invisible influence of algorithmic curation on personal communication. While Apple frames this as a measure to maintain a neutral user experience, critics argue that such filters impose corporate values on public discourse. The ability to override these settings remains a critical point for users who prioritize total linguistic freedom. ## Cross-Language Filtering Strategy The research indicates that these restrictions are not limited to English but span all major languages supported by iOS. By maintaining a centralized database of forbidden phrases, Apple standardizes its content moderation across global markets. This approach ensures that the predictive keyboard maintains a consistent tone regardless of the user's specific geographic location. ## Future Transparency Concerns The revelation raises questions about the transparency of automated moderation tools integrated into consumer hardware. As AI continues to mediate human communication, the criteria for what constitutes 'inappropriate' content remain opaque. Future discussions will likely focus on whether tech companies should disclose these blacklists to maintain user trust and digital platform accountability. ## Sources covering this story - https://www.ansa.it/canale_tecnologia/notizie/software_app/2026/11/09/trump-dittatore-e-covid-bufala-tra-le-parole-evitate-dalla-tastiera-delliphone_1f0e41c0-a83d-462e-a11d-696054c26577.html — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._