A threat against a judge made in a conversation with AI reaches the courts: chats are now evidence

By malvern-online.com · November 5, 2026.

According to a malvern-online.com headline, a man described as a sex offender is accused of threatening a circuit judge during a conversation with an artificial intelligence. He remains in custody on $1 million bail. We only have the headline and a brief description. We have not been able to verify the article's content, so we do not know which chatbot was involved, the exact wording of the alleged threat or how it came to light. This is an accusation, not a conviction. For now, none of what follows is established fact.

Our take is that what matters here is not the chatbot but the shift in context. Many people treat these conversations as a private space, almost a diary. A chat with an AI, however, leaves a record that can end up before the authorities. In general, AI services retain histories and can be served with court orders. Anyone who assumes that what they tell an assistant stays on the screen is mistaken.

Cases like this raise two tensions we will be following closely. The first is privacy: what is retained, who reviews it and when it is reported to the authorities. The second is proportionality: a sentence typed to a machine is not the same as an actionable threat, and courts will have to distinguish between venting and genuine intent. In the short term there will be more episodes like this, with both hits and misses. In the long term, clear, evidence-based rules on how these conversations are handled need to be established. Those rules must protect victims without turning every chat into a case file.