Alphabet’s Google has started training a version of its generative-AI software that can operate directly on Apple’s servers, Bloomberg reported on 22 August. The move would allow Apple to use Google technology for a revamped Siri and other consumer features while keeping data on its own infrastructure, marking a rare deep technical collaboration between two long-standing rivals in mobile computing. The Apple work follows a separate agreement disclosed the same day in which Meta Platforms committed to spend more than $10 billion over six years on Google Cloud’s infrastructure, including its seventh-generation Tensor Processing Units and Vertex AI platform, according to industry newsletter The Rundown Tech. Meta said the capacity will supplement its in-house data-centre fleet as it scales generative-AI services. Taken together, the prospective Apple deployment and the confirmed Meta contract underscore Google’s strategy of licensing its AI stack and selling cloud capacity to outside tech giants as investment in artificial intelligence accelerates. Investor reaction has been positive: Alphabet shares reached a record high during Thursday trading after the reports surfaced.