Jacob Coxon, who said he spent three years doing research at both Anthropic and OpenAI, said Tuesday on the social platform X that the two AI companies are more focused on beating each other and global competitors in developing the most advanced model possible than they are on safety.