Security & Misuse

An article examining operations and hacking techniques related to content metadata and time information using C2PA technology.

Apple announced plans to introduce additional controls for granting full disk access on macOS to mitigate data leakage risks from AI agents.

• •Stricter permission-granting procedures are being implemented to reduce data leakage risks associated with autonomous AI agents.
• •Explicit user actions and additional controls are being considered for granting full disk access.
• •The move follows discussions regarding message data monitoring related to Meta's AI agent, Muse.

Apple modified full disk access permissions to curb AI agent misuse, while Meta claims FDA is insufficient for Muse to read messages, a stance Apple disputes.

• •Meta stated that Full Disk Access (FDA) is not sufficient for Muse to read messages.
• •Apple disputed Meta's claim.

Apple is introducing new restrictions and management features for full disk access on Mac to address risks associated with AI agents.

• •As previously reported by TechCrunch, Apple is adding new restrictions to full disk access on Mac to address risks brought by AI agents.
• •In a Friday update, Apple stated it is deploying new controls to ensure only users who genuinely want to grant this high-level access can do so.

Apple warned of risks from advanced AI agents and announced new management features for Full Disk Access permissions in macOS.

• •Apple plans to add new management features surrounding Full Disk Access permissions in macOS.
• •Apple warned that capable AI agents pose risks by gaining broad access to user files, messages, emails, and browsing history.
• •The changes aim to enhance safety when AI agents access user data.

Google Research is investigating methods to achieve provably private learning from federated data.

A New York server named Madison reports that AI hallucinations are encouraging customer entitlement and complicating on-site service.

• •Madison, a server at a New York restaurant, is facing troubles caused by misinformation from AI.
• •Cases are increasing where customers receive incorrect information from AI, such as allergy misunderstandings, and skip on-site verification.
• •AI hallucinations are worsening customer demands and increasing the burden on service providers.

Latest releases of FortiNDR Cloud and FortiNDR were announced, enhancing features to counter AI-driven attacks and prepare for quantum-era threats.

• •Latest releases are provided to address attacks that leverage AI to amplify activities.
• •New features include enhanced quantum readiness and automated deception.
• •Features enabling on-premises AI deployment are included.

Video content by Greg Kroah-Hartman examining security challenges and impacts arising from the proliferation of large language models.

This article explains how endpoints, now handling inference in the AI era, face diverse threats as targets of intensifying cyberattacks.

• •Endpoints have begun taking on the new role of inference processing in the AI era.
• •They are exposed to diverse threats as targets of intensifying cyberattacks.
• •In an era assuming breach, the true value of endpoints is being questioned.