Artificial intelligence has been widely predicted to disrupt smartphones, and hurt Apple, but Perplexity CEO Aravind Srinivas argues that the iPhone will become more important as AI improve
In a segment from "This Week in AI" published on April 23, Aravind Srinivas, co-founder and CEO of AI search company Perplexity, lays out an outside view of Apple's position. Srinivas, who previously worked in AI research roles at OpenAI, Google, and DeepMind, argues that AI may reinforce Apple's core products instead of replacing them.
"The phone, the iPhone is actually not getting disrupted by AI at all," Srinivas says, arguing that better AI pushes the device in the opposite direction, turning it into "your digital passport."
The claim runs against the idea that Apple is falling behind in AI compared to companies like OpenAI and Google. Srinivas acknowledges the gap, but he shifts the focus to hardware, privacy, and user data control.
The iPhone's role in a more personal AI era
Srinivas frames the iPhone as the place where personal context already lives. Payment methods, identity credentials, health data, communications, and photos all sit on a single device tied to one user over time.
AI systems increasingly rely on context to produce useful results, a shift already visible in tools like Perplexity's Personal Computer feature.
"All these are things that are truly personal to you," Srinivas mentions, pointing to a layer Apple already controls and can build on.
His framing positions the iPhone less as a product at risk of disruption and more as infrastructure for future AI systems. Greater AI capability increases the value of personal data, which increases the importance of the device that holds it.
Apple Silicon and the shift toward on-device intelligence
The argument leans heavily on Apple's hardware strategy. Srinivas calls Apple Silicon "one of their underrated assets," especially as AI workloads begin to move closer to the user.
He describes a shift toward AI systems that run closer to the user instead of relying on centralized infrastructure. "If agent loops start running locally, that doesn't need to be centralized on servers," Srinivas explains, with those workflows tied to "local files, local apps, messages, emails, notes, photos."
On-device processing aligns with Apple's long-standing privacy model. Keeping sensitive data local reduces exposure and avoids sending personal information to external servers.
Apple holds a structural advantage in its control over hardware, software, and personal data. No other company combines those layers at the same scale within a single ecosystem.
However, Apple hasn't demonstrated a frontier model that directly competes with OpenAI or Google. Siri still falls behind newer conversational systems that have already set the standard for user expectations.
Users already rely on third-party AI tools on Apple devices, which shows that control of hardware doesn't guarantee control of intelligence. Apple owns the platform, but competitors still define much of the AI experience running on it.
