Apple is reportedly building its own AI-powered answer engine, aiming to reduce its dependence on partners like OpenAI and Google. According to Bloomberg’s Mark Gurman, a new internal team—called “Answers, Knowledge, and Information”—is working on a system designed to fetch and deliver real-time information from the web, similar to ChatGPT or Perplexity. The project could show up as a standalone app or be integrated into Siri, Safari, and Spotlight.
Apple has been slow to roll out meaningful AI upgrades. Its Siri revamp isn’t expected until 2026, and internally, the company admits it’s at least two years behind major players in generative AI. In the meantime, Apple has leaned on OpenAI’s models to power some of its AI features in iOS 18. But long term, it seems determined to build a deeper, more self-reliant stack.
The company is hiring engineers with backgrounds in search engine design, and the new system would likely build on Apple’s existing AI infrastructure—like its on-device and server-based models running under Apple Intelligence. Earlier leaks have revealed internal tools like Ajax (a framework for training models) and a prototype chatbot dubbed “Apple GPT,” though nothing public has come of them yet.
Apple’s broader AI strategy also ties into its uneasy relationship with Google. The two companies have a default search agreement reportedly worth $20 billion a year, but Apple is already exploring adding AI-based alternatives like Perplexity to Safari. If Apple successfully launches its own engine, it could not only modernize Siri and Spotlight, but also threaten Google’s grip on iPhone search.
The push reflects Apple’s need to catch up fast while maintaining its usual emphasis on privacy and control. Whether this answer engine ships anytime soon—or stays an internal project—could shape Apple’s entire approach to AI going forward.