Perplexity Launches Personal Computer for Mac
Perplexity has launched Personal Computer for Mac, a system that acts as a persistent AI agent, integrating with local files, native applications, and browsers to execute complex workflows autonomously for its premium subscribers.
The News
On April 16, 2026, Perplexity began rolling out its 'Personal Computer' feature to all Max subscribers and waitlist members. This new system, available through an upgraded Mac desktop application, expands on the web-based 'Perplexity Computer' by integrating directly with the user's local machine. Personal Computer can access and edit local files, interact with native Mac applications like iMessage and Calendar, and orchestrate complex workflows across them 24/7, particularly when installed on a dedicated machine like a Mac mini. The feature is positioned as a premium offering, available to $200/month Max subscribers.
The OPTYX Analysis
The launch of Personal Computer marks a significant step beyond conversational AI toward agentic computing. Instead of acting as a passive question-answering tool, Perplexity is positioning its AI as an active, autonomous 'digital worker' that can execute tasks persistently in the background. By integrating with the local file system and native applications, it moves from processing public web data to managing and acting upon a user's private information and workflows. This strategy aims to create deep user lock-in by embedding the AI into the core, day-to-day operational fabric of the user, transforming the AI from a search tool into a personalized operating system layer.
AI Control Impact
Enterprise adoption of personal AI agents introduces a new vector of data governance and security risk. The primary vulnerability is the unmonitored execution of tasks by an AI agent with broad access to sensitive local files, internal applications, and communication channels. An improperly configured or compromised agent could lead to data exfiltration, unauthorized actions, or compliance violations. The immediate operational fix requires CIOs and CISOs to establish clear policies on the use of such persistent AI agents on corporate devices. This includes defining permissible levels of access, implementing sandboxing where possible, and requiring auditable logs of all AI-initiated actions to ensure that agent autonomy does not create an ungoverned, high-risk attack surface.