AI··2 min readTechnical readSource-backed

Apple Redesigns Mac Security to Keep Autonomous AI Agents on a Leash

Autonomous software tools that take actions across your operating system bring massive privacy risks. Unlike standard chatbots that just answer questions, these local helpers can read files and control apps on your behalf.

By
WIREUNWIRED BRIEF

Autonomous software tools that take actions across your operating system bring massive privacy risks. Unlike standard chatbots that just answer questions, these local helpers can read files and control apps on your behalf. Apple is responding to this danger with updated security controls in macOS 27.3. The company warns that granting an app full disk access puts all personal data on your machine at risk.

Two recent security scares forced Apple's hand. A journalist reported that Meta’s Muse app read private messages without consent, while a separate Wired report revealed a ChatGPT Mac app flaw that could expose sensitive data. These incidents highlight the threat posed by apps demanding sweeping system privileges.

Yet, Apple also benefits financially when early adopters buy powerful Mac minis and Mac Studios to run these exact local models. The upcoming software update splits Device Control and Data Access into two distinct settings: System & App Protection Override and Accessibility. This separation forces users to explicitly acknowledge they are overriding built-in security protections.

Furthermore, a new granular control option lets you configure apps to ask for permission every time they request access.

While this manual approval process creates an obstacle to true automation, it stops agents from getting carte blanche over your entire hard drive.

Our read: this friction between automation and safety threatens the very frictionless promise of AI agents.

If users must constantly click approve, the workflow breaks down.

That tension suggests developers must build apps that gracefully handle restricted folders, rather than demanding total system access out of the box.

STORY JOURNEY

The story so far

Earlier coverage connected by shared names and topics. Follow the dates to see how coverage developed; these links do not imply every article describes the same event.

  1. Microsoft Introduces New Security Tools and Guardian Agents to Mitigate Risks from Unpredictable Corporate AI

    Microsoft introduced new security tools on Monday to protect businesses from the growing risks associated with corporate artificial intelligence agents. The company…

    Connected by: SecurityEARLIER COVERAGE
  2. Editorial illustration for DoorDash Reality Check Shows AI Agents Are Still Generating Mere Hundreds of Orders
    DoorDash Reality Check Shows AI Agents Are Still Generating Mere Hundreds of Orders

    Silicon Valley loves to talk about autonomous AI agents taking over human commerce, but reality moves at a much slower pace. Delivery…

    Connected by: AgentsEARLIER COVERAGE
  3. Editorial illustration for Apple Pushes Back Against the Rush to Ship AI
    Apple Pushes Back Against the Rush to Ship AI

    Apple wants you to stop thinking about artificial intelligence altogether. During a recent interview with Wired, newly appointed CEO John Ternus pushed…

    Connected by: AppleEARLIER COVERAGE
  4. Editorial illustration for Meta’s AI agent Muse lands on Apple Watch and Android Wear OS
    Meta's Muse Lands on Your Wrist, Chasing the Ultimate AI Habit

    Your wrist is the next battleground for artificial intelligence. Meta is bringing its Muse AI agent to the Apple Watch and Android's…

    Connected by: AppleEARLIER COVERAGE
  5. NOW
    Apple Redesigns Mac Security to Keep Autonomous AI Agents on a LeashCURRENT UPDATE
A WireUnwired summary is already available. You can still write your own brief below for the reader community.
SOURCE & ATTRIBUTION
TechCrunch

Original author: Sarah Perez

Original publication: 17 November 2026 02:31

Read full story on TechCrunch ↗
RELATED STORIES
TechCrunchCrowdStrike Bets Nearly Half a Billion on the Next Security FrontierEditorial illustration for LangChain Doubles Valuation to $3B on the Back of Enterprise AI AgentsTechCrunchLangChain Doubles Valuation to $3B on the Back of Enterprise AI AgentsEditorial illustration for OpenAI gives ChatGPT a wallet so AI agents can spend your moneyTechCrunchOpenAI gives ChatGPT a wallet so AI agents can spend your money
Original sourceTechCrunch ↗