When will Apple and OpenAI’s new partnership start generating revenue for either company?
According to a report Wednesday (June 12) from Bloomberg News, not immediately.
As reported earlier this week, Apple announced that it would incorporate OpenAI’s AI model ChatGPT into its Siri digital assistant, with ChatGPT4o added to iOS, macOS, and iPadOS, “ready to tag in when Siri taps out.”
Sources close to the deal told Bloomberg that, instead of a payment to OpenAI, Apple sees the exposure of OpenAI’s brand and product to millions of devices as a valuable alternative to a monetary compensation.
In return, Apple can provide users with a more advanced chatbot, potentially increasing user engagement and driving upgrades.
The report mentions that, currently, the partnership could be costly for OpenAI, as they have to pay Microsoft to host ChatGPT on their cloud network. With more ChatGPT users, OpenAI’s expenses increase, especially with the integration into iPhones posing a potential budget challenge.
Although ChatGPT will be offered for free on Apple devices, revenue can still be generated by converting free users into paid subscribers. Any transactions from users subscribing to OpenAI on an Apple device via the ChatGPT app will go through Apple’s payment platform, allowing the tech giant to earn a percentage.
PYMNTS previously stated this week that the partnership between Apple and OpenAI resembles the collaborations seen among banks and FinTechs.
“The classic dilemma of whether to build an in-house solution, buy a ready-made product, or form a partnership for integrating new technologies has long been a focal point of business development. The need to partner with third-party vendors has become increasingly important,” as mentioned by PYMNTS.
Industry experts also discussed the Private Cloud Compute service within Apple’s AI ecosystem.
Yannik Schrade, CEO and co-founder of computing startup Arcium, applauded the service as a positive move for Apple but emphasized that more steps are needed.
“Trusted hardware-based confidential computing has been around for some time and faces security challenges, necessitating trust in third parties that ideally should not be required from users,” Schrade explained.