How to recognise AI-written text? ‘Countermeasures’ exist, but it’s a long way to go – Daniela De Pasquale’s in-depth analysis
On Agenda Digitale, an in-depth analysis by our Partner Daniela De Pasquale about ChatGPT and generative AI detection systems.
Given the high-speed development of generative AI, some key issues remain unresolved to date, including the need to understand whether a text, work, diagnosis, research or judgment was written entirely or partially with an AI-originated text, or whether students can make use of it during written exams, distorting the results.
Hence, in recent weeks, the rush by industry to offer technological solutions answering the question of whether a certain text was generated by human beings or AI.
But how effective are these detection tools?