Videos shared online, external show people speaking the word “racist” into the Dictation tool.

Sometimes it is transcribed correctly – but on other occasions it is turned into “Trump”, before being quickly restored to the correct word.

The BBC has not been able to replicate the mistake, suggesting Apple’s fix is already taking effect.

Prof Bell said Apple’s explanation of phonetic overlap did not make sense because the two words were not similar enough to confuse an artificial intelligence (AI) system.

Speech-to-text recognition models are trained by inputting clips of real people speaking alongside an accurate transcript of what they say.

They are also taught to understand words in context – for example, they could distinguish the word “cup” from “cut” if it was within the phrase “a cup of tea”.

Prof Bell says the situation with Apple is unlikely to be a genuine mistake with its data because its English language model would be trained on hundreds of thousands of hours of speech, which should give it a high level of accuracy.

For “less well-resourced languages” he said it could be an AI training issue.

But he said in this case: “it probably points to somebody that’s got access to the process.”

A former Apple employee who worked on its AI assistant Siri told the New York Times:, external “This smells like a serious prank.”