AI systems are learning to lie and deceive, scientists find

Research has revealed concerning findings about AI systems known as large language models (LLMs) and their ability to deceive human observers intentionally.