(NewsNation) — A 60-year-old man wound up in the hospital after seeking dietary advice from ChatGPT and accidentally poisoning himself.

According to a report published last week in the Annals of Internal Medicine, the man wanted to eliminate salt from his diet and asked ChatGPT for a replacement.

The AI platform recommended sodium bromide, a chemical often used in pesticides, as a substitute. The man then purchased the sodium bromide online and replaced it with salt for three months.

The man eventually went to the hospital, fearing his neighbor was trying to poison him. Doctors discovered he was suffering from bromide toxicity, which caused paranoia and hallucinations.

Bromide toxicity was more common in the 20th century when bromide salts were used in various over-the-counter medications. Cases declined sharply after the U.S. Food and Drug Administration phased out bromide between 1975 and 1989.

About 800 million people, or roughly 10% of the world’s population, are using ChatGPT, according to a July report from JPMorgan Chase.

“It’s technology that has the potential to enable enormous leaps in productivity and human understanding,” said Imran Ahmed, CEO at the Center for Countering Digital Hate, a watchdog group. “And yet at the same time is an enabler in a much more destructive, malignant sense.”

New research from the group that focused on teens revealed that ChatGPT can provide harmful advice.

The Associated Press reviewed interactions where the chatbot gave detailed plans for drug use, eating disorders, and even suicide notes.

The Center for Countering Digital Hate found that testing ChatGPT with harmful prompts led the chatbot to respond in dangerous ways more than half the time.

The study highlights the risks as more people turn to AI for companionship and advice. OpenAI, the maker of ChatGPT, said after viewing the findings that its “work is ongoing” in refining “how models identify and respond appropriately in sensitive situations.”

The Associated Press contributed to this report.