A year after our initial research on threat actors’ attitudes to generative AI, we revisit some underground forums and find that many cybercriminals are still skeptical – although there has been a slight shift