Apple’s new iPhones could boost AI usage inside companies. But CIOs have questions.

Apple on Monday introduced a new generation of iPhones that put AI at the forefront of the device’s operating system, a test of consumer appetites for such tools. The new tools—called “Apple Intelligence"—include an improved Siri voice assistant and a variety of text-generation and photo-editing capabilities, and they will be compatible with the new iPhone 16 and the iPhone 15 Pro. Apple announced most of the AI functions in June. A beta, or early version, of some of the Apple Intelligence features will be available next month.

The company’s shares fell by more than 1% during Monday’s presentation, reflecting the market’s skepticism of AI-centric products. But for enterprises that remain bullish on AI’s potential, the devices represent a new era of possibilities and concerns.

Shawn Malhotra, chief technology officer of mortgage originator and financial services company Rocket Cos., said he is excited to let employees experiment with on-device AI, which allows users to run AI algorithms without the need for a remote cloud server. “It would be a mistake to prevent people from using new technologies," he said.

Still, Rocket needs to make sure its data isn’t at risk, and that any usage meets information security requirements.

The new iPhone line includes a chip that Apple executives say will help power the new AI features.

Kunal Anand, chief technology and AI officer of application security and delivery company F5, has been running a beta test of Apple Intelligence since August.

So far, he’s seen less clarity and transparency and fewer systems for monitoring exactly when and where AI apps will be running on the device itself versus on the cloud—where CIOs typically want to avoid having sensitive data.

“It’s extremely unclear," said Anand.

Apple said that for Apple Intelligence, which runs on Apple’s proprietary AI models, users will be able to access a transparency log that shows them which requests were processed on device versus which were processed on Apple’s Private Cloud Compute, its cloud intelligence system designed specifically for private AI processing.

The company also said that it has strict security controls around its Private Cloud Compute setup, so that data requests processed on its Private Cloud Compute essentially have the same level of security as those processed on device.

It’s a slightly different story when it comes to ChatGPT, the company said. Apple earlier announced a deal with OpenAI, where the startup’s chatbot would handle some functions that Apple’s own AI can’t. For some features, such as full-out text generation, the phone might suggest that it’s a request better handled by ChatGPT, and then ask the user’s permission before sending it out to ChatGPT’s servers. Apple also said that this feature will be off by default, and require users to actively turn it on.

Additionally, Apple has said that the devices include systems that allow IT departments to manage users’ access to Apple Intelligence features, including the ChatGPT integration.

Anshu Bhardwaj, senior vice president and chief operating officer of retailer Walmart’s global technology unit, said both enterprises and tech providers are motivated to work together to embed the controls required to get these devices to the point where they’re enterprise ready. Enterprises—not consumers—are the ones that will ultimately drive the use of AI at scale, Bhardwaj said.

“Once the device manufacturers figure out those things from a data privacy security standpoint, then I feel like you’ll see the adoption go faster," she said.

Bhardwaj added there is significant potential for on-device AI at Walmart, where on the floor associates are already equipped with Samsung devices.

Ultimately, getting AI into employees’ hands from a consumer perspective could also drive greater adoption on the enterprise side, said Christian Frank, technology director of credit-rating provider S&P Global Ratings.

“Apple is going to do a lot of work here with consumers," Frank said. “They’re very good at customer experience, so I feel like they’re going to figure out some use cases to get into consumers’ hands that will perform well" and that would also benefit the enterprise.

One of the biggest stumbling blocks to AI adoption in enterprises right now is just the lack of familiarity with the technology as well as what it can do, how to use it and what use cases are best.

“The smartphone players, I think, can help broaden the number of those use cases," he said.

Write to Isabelle Bousquette at isabelle.bousquette@wsj.com