Google will save your Lens photos, Search Live recordings, and Translate audio for AI training

Editor's take

Google is now collecting user data from Lens, Live, and Translate for AI training, consolidating it under a new "Search Services History" setting. This move signals a deeper integration of user interaction data into Google's foundational AI models, aiming to refine their understanding and capabilities across visual search, real-time translation, and potentially other multimodal AI applications.

This broadens Google's data moat, a critical asset in the generative AI race against competitors like OpenAI and Microsoft, who are also heavily reliant on vast datasets. Users who previously opted out of specific data collection might find their interactions now indirectly contributing to AI development, raising privacy considerations and the need for transparent opt-out mechanisms beyond the new default.

Future developments to monitor include the specific types of AI models benefiting from this aggregated data and whether Google will offer granular controls over which specific service data is used for training. The impact on the quality and bias of future Google AI products, particularly as they become more multimodal and context-aware, will be a key indicator of the success and implications of this data strategy.