But Google and its hardware partners say privacy and security are the primary focus of Android AI’s approach. Vice President Justin Choi, head of the security team, mobile eXperience business at Samsung Electronics, says its hybrid AI offers users “control over their data and uncompromising privacy.”
Choi describes how cloud-based features are protected by servers that are subject to strict policies. “Our on-device AI features provide another layer of security by executing tasks locally on the device, without relying on cloud servers, and neither storing data on the device nor sending it to the cloud,” Choi says.
Google says its data centers are designed with resilient security measures, including physical security, access controls, and data encryption. While AI requests are processed in the cloud, the company says the data remains within Google’s secure data center architecture, and the company does not send your information to third parties.
Meanwhile, Galaxy’s AI engines aren’t trained on user data from on-device features, Choi says. Samsung “clearly indicates” which AI features are running on a device with the Galaxy AI symbol, and the smartphone maker adds a watermark to show when content uses generative AI.
The company also introduced a novel security and privacy option called Advanced Intelligence Settings to give users the ability to disable cloud-based AI features.
Google says it “has a long history of protecting the privacy of user data,” adding that this applies to AI features powered on devices and in the cloud. “We use on-device models where data never leaves the phone for sensitive use cases like phone call filtering,” Suzanne Frey, Google’s vice president of product trust, tells WIRED.
Frey describes how Google products rely on cloud-based models that she says ensure that “consumer information, such as the sensitive information you want to summarize, will never be sent to a third party for processing.”
“We remain committed to building AI-powered features that people can trust because they are secure by default, designed with privacy in mind, and most importantly, they adhere to Google’s Responsible AI Principles, which were an industry first,” Frey says.
Apple changes the conversation
Rather than simply adapting a “hybrid” approach to computing, experts say Apple’s AI strategy has changed the nature of the conversation. “Everyone expected this push into devices, putting privacy first, but what Apple has really done is say it doesn’t matter what you do in AI—or where—it matters how you do it,” Doffman says. He believes it “will probably define best practices in the smartphone AI space.”
Still, Apple hasn’t won the AI privacy battle yet: The OpenAI deal — in which Apple, uncharacteristically, hands over its iOS ecosystem to a third-party vendor — could put an end to its privacy claims.
Apple rejects Musk’s claims that the OpenAI partnership compromises iPhone security because “users who access ChatGPT have built-in privacy protections.” The company says you’ll be asked for permission before your query is shared with ChatGPT, while IP addresses are hidden and OpenAI won’t store requests — but ChatGPT’s data usage policies still apply.
Partnering with another company is a “strange move” for Apple, but the decision “wouldn’t be taken lightly,” says Jake Moore, global cybersecurity advisor at security firm ESET. While the exact privacy implications aren’t yet clear, he acknowledges that “some personal data could be collected on both sides and potentially analyzed by OpenAI.”