Apple CEO Tim Cook recently acknowledged DeepSeek’s AI models as a step forward in efficiency while addressing analysts’ questions during an earnings call.
He discussed Apple’s AI strategy, emphasizing the company’s hybrid model, which runs simpler AI tasks locally while handling more complex ones in the cloud through partnerships with AI providers.
Currently, Apple’s only AI partnership is with OpenAI, allowing ChatGPT to assist iPhone users through a private cloud.
Cook’s comments come amid OpenAI’s claim that DeepSeek used its models for training through a method called distillation, which would violate OpenAI’s policies.
Despite these allegations, Cook still commended DeepSeek’s AI development, though he did not suggest Apple would integrate its models into the iPhone.
Apple has indicated that its partnership with OpenAI is not exclusive and may bring in other AI models in the future, such as Google’s Gemini or Anthropic’s Claude.
Beyond OpenAI’s accusations, industry analysts have questioned the efficiency of DeepSeek’s training process, suggesting it required more GPUs and computational power than the company stated. However, DeepSeek continues to offer its AI models at significantly lower prices than its American competitors.
During the call, Cook also addressed Apple Intelligence and its phased rollout. While some expected a boost in iPhone sales due to new AI features, sales declined slightly compared to the previous year. Cook noted that regions, where Apple Intelligence was available, saw stronger sales than those without it.
Apple has also faced challenges with its AI features. Recently, it paused its AI-generated news summaries after inaccuracies, including a false headline about a high-profile crime case. The company is now working to refine its AI capabilities as it continues expanding its AI-driven offerings.
