According to The Information on August 17, Anthropic and OpenAI, while accelerating their deployment of vertical industry applications, may tighten their API access, which creates direct competition with their own API clients and raises concerns among enterprise customers. The reasons behind this move include pressure from the Donald Trump administration on model security, preventing model distillation, and the potentially more lucrative returns from vertical applications. Anthropic's Q2 revenue has already reached $11.5 billion. Facing this uncertainty, leading developers such as legal AI company Harvey have begun training their own models to reduce their reliance.