At least 50 per cent of teachers have had to mark assignments they were certain had been written by ChatGPT or other generative AI platforms, but their schools had no policies to deal with the problem, a government inquiry into the use of artificial intelligence in education has heard.

That issue is compounded by there being no tools available to help teachers detect AI-generated content, and even linguists say it is impossible to tell human- and computer-generated text apart, a panel of four education experts told the House standing committee on employment, education and training on its first day of public hearings.