Apple plans to start analyzing data on users’ devices to improve its artificial intelligence platform. The move is said to protect user information while helping the company catch up with competitors in the AI field.
Apple currently trains AI models on synthetic data that only mimics real data and does not contain any user information. However, synthetic data does not always match real customer information, which ultimately makes it difficult for AI algorithms trained on this data to work correctly.
The new approach will address this issue while ensuring that user data remains on customers’ devices and is not directly used to train AI models. The main goal is to help Apple catch up with competitors like OpenAI and Alphabet, which have fewer restrictions on customer data privacy.
The technology works like this: An algorithm takes Apple-generated synthetic data and compares it to a live sample of user emails in the company’s Mail app for iPhone, iPad, and Mac. By using real emails to check the accuracy of the generated synthetic data, Apple will have a better understanding of which parts of the synthetic data are most representative of real messages and therefore suitable for training AI models. This data will help the company improve the AI functions associated with text generation in the tools that make up the Apple Intelligence platform.
«By generating synthetic data, we aim to produce synthetic sentences or emails that are similar enough in topic or style to real ones to improve our models for generalization, but without collecting emails from Apple users’ devices,” the developers said in a statement.
In addition to synthetic data, Apple trains its AI models on information it licenses from third parties or collects publicly from the internet. Using synthetic data has its own drawbacks, which means Apple’s algorithms often don’t do a great job of performing tasks like text summarization.
In theory, the new approach will help improve Apple’s AI models, which would be a key step toward competing with the leaders in the AI segment. Apple acknowledges that its products lag behind the market leaders and is working to improve the situation. “By drawing on years of experience using techniques like differential privacy, as well as new technologies like synthetic data generation, we can improve Apple Intelligence features while protecting the privacy of users who have opted in to participate in the device analytics program,” Apple said in a statement.