The UK’s AI Security Institute (AISI) was not offered access to Anthropic’s latest large language model, Claude Mythos 5.1, according to the FT. The institute, described internally as “a start-up in the government”, had previously been offered unique access to pre-deployment models through a system of patronage dependent on the good grace of American frontier companies and, officials suspect, the US government. In the same week, Anthropic researcher Jacob Coxon resigned, writing that he was deeply concerned about AI technology wiping out humanity; fellow Anthropic employee Evan Hubinger agreed with Coxon’s concerns and said there was a 10% chance of AI killing all humans. The article states there is legitimate concern among both the public and the British political class about these developments, but that concerns surrounding AI safety are being misdirected in ways likely to make us more vulnerable.