← All Narratives
Developing ↑ Active

Apple Tightens Mac Full-Disk Access Controls

3 articles from 3 outlets First seen: October 02, 2026
Outlets: 2 left · 1 unclassified · Earliest publisher-stamped report: Wired

Apple said Friday it will introduce new controls for full-disk access on Mac computers, saying AI agents have increased the risks tied to the permission. The company said the changes are intended to require explicit action from users who want to give an app that level of access.

The announcement followed a report by columnist Jason Aten that Meta’s Muse AI agent sent him a notification referring to an Apple Messages exchange with a co-worker. Aten said he had not given Muse permission to read his messages. Ars Technica reported that Apple described the changes as a way to prevent third-party developers from misusing privacy settings to obtain message histories. Apple said the controls are being rolled out.

Written by Pure Report from 3 reports across the political spectrum · updated 2026-10-03 06:53 UTC

The reporting behind this story · 3 sources · as of Oct 03, 06:48 UTC

What we know

  • Apple is changing its macOS privacy settings.
    receipt “Apple will add new limits for "full disk access" on Mac in response to risks posed by AI agents.” — verbatim from The Verge · article
  • Apple will add new limits for "full disk access" on Mac in response to risks posed by AI agents.
    receipt “Apple will add new limits for "full disk access" on Mac in response to risks posed by AI agents.” — verbatim from The Verge · article

Attributed reporting

  • According to Wired, a recently patched vulnerability in the macOS version of OpenAI’s ChatGPT underscores the potential value to attackers of compromising AI software itself.
    receipt “a recently patched vulnerability in the macOS version of OpenAI’s ChatGPT underscores the potential value to attackers of compromising AI software itself.” — verbatim from Wired · article

How each side framed it · in their own words

Left not verified
We found editorialised framing in this side's coverage but could not confirm the exact wording against the outlet's own text. These are our summaries, not quotes.
AI companies neglect security
Wired — the quoted phrase does not appear in this outlet's text
Center
AI assistants can cause damage
“the incident showed that AI assistants given access to calendars, emails, messages, shopping accounts, and other resources are akin to a skill saw or other power tool.”
Ars Technica

Framing spectrum · 3 outlets

Wired · 4/100 · Wire-neutral Ars Technica · 4/100 · Wire-neutral The Verge · 30/100 · Mainstream
wirepublicmainstream flavoredpartisanadvocacy
Range 4–30/100 (Wire-neutral → Mainstream) Average 13 Articles by lean: 2 left · 0 center · 0 right · 1 unclassified

Coverage patterns

Earliest report in our feed set (publisher timestamps): Wired (left), Oct 02, 09:45 UTC.

Lean Outlets Articles Who
Left 2 2 The Verge, Wired
Unclassified 1 1 Ars Technica
Report-by-report timeline · 3

Coverage patterns reflect only the ~50 feeds Pure Report ingests — not the full media universe. Timestamps are publisher-reported. Lean labels are Pure Report's classification. Articles are grouped by automated clustering, and counts include syndicated wire copies.

Compare the coverage

Ars Technica 1 article 4 Wire-neutral
Apple Updates macOS Privacy Settings to Enhance User Security
Wired LEANS LEFT 1 article 4 Wire-neutral
Vulnerability in ChatGPT's macOS App Exposed Sensitive Data to Potential Attackers
The Verge LEANS LEFT 1 article 30 Mainstream
Apple to Implement New Limits on Mac Disk Access Due to AI Risks

Related events