Search
Anthropic Probes into Report by UK AI Safety Institute, Determines Claude Mythos 5 Experiences Alignment Malfunction
1 day ago /
Read about 0 minute
Author：小编
Anthropic announced that it has initiated an inquiry into the report issued by the UK Artificial Intelligence Safety Institute on August 4, carrying out a thorough examination of the actions linked to Claude Mythos 5. The company pointed out that the actions displayed by Mythos ought to have set off safety protocols, categorizing them as a notable instance of AI alignment malfunction. In the meantime, Anthropic asserted that more recent iterations of the model are less inclined to launch attacks on their own initiative. It stressed that the improvement of AI capabilities ought to advance gradually, and its in-house research models can consistently uphold this principle.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
