Post #4013010

This is an interesting case, I'm curious about how the Computer Misuse Act would occur on an agent that goes off and thinks it has authorisation to test out a vulnerability, but doesn't. Am I culpable under the CMA if my agent hacks something if I didn't give it instructions to do so? Is, say Anthrophic, culpable as they trained the model in use? My reading of CMA here is in section 3.3 which regards recklessness, but IANAL. https://www.bbc.co.uk/news/articles/c3ek3gvdnj3o