Apple fixing iPhone bug that mistakenly typed 'Trump' instead of 'racist'

Apple is fixing a bug in the iPhone's dictation feature after users noticed it sometimes mistakenly transcribed the word 'racist' as 'Trump'. The issue gained attention on TikTok and other social media platforms.  

Apple acknowledged the problem and said it was caused by a flaw in the speech recognition model. The company has already started rolling out a fix.  

Users reported that when they said 'racist' while using the dictation feature, the iPhone would briefly display 'Trump' before correcting itself. According to an NYT report, Apple explained that the speech recognition system sometimes shows words that sound similar before settling on the correct one.  

Similar controversy with Amazon's Alexa

This isn't the first time a tech company's voice recognition system has caused political controversy. Before the 2024 U.S. election, Amazon’s Alexa gave different responses when asked why someone should vote for Donald Trump versus Kamala Harris. Alexa refused to provide reasons for voting for Trump, while in some cases, it listed reasons for voting for Harris. Amazon later called this an error and fixed it.  

Separately, Apple announced plans to invest over $500 billion in the U.S. over the next four years. This includes hiring 20,000 employees and producing Apple TV+ content in 20 states. The announcement came after Apple CEO Tim Cook met with former President Trump last week, though it's unclear if the investment plan was influenced by that meeting.

Scott Stephenson, Founder and CEO of Deepgram (Deepgram’s voice AI platform provides APIs for speech-to-text, text-to-speech, and full speech-to-speech voice agents), said, "AI is only as smart as the data it’s trained on. Voice recognition should be about understanding, not assuming. This is a reminder that companies need to constantly refine their models to avoid bizarre and potentially harmful mix-ups. The goal isn’t just accuracy, it’s trust."

He explained that if AI can’t understand what we’re saying by accommodating accents, slang, and emotions, then it’s just noise. The next wave of innovation is about making AI listen smarter, not just faster.