No, absolutely not! I don’t trust Apple for the same reasons I don’t trust other big tech corporations or any closed-source ecosystem.

If your trust comes from “they promised they’re not evil,” mine comes from documented behavior that says otherwise. Their reputation relies on marketing slogans like “we protect your privacy” while my skepticism is based on facts, history and hard evidence.

“Client-side scanning” (CSS) refers to analyzing data, photos, messages, files, locally on your device before it’s encrypted or sent. The idea is that AI or machine learning models scan your content for illegal or harmful material (like Child Sexual Abuse Material, CSAM) without Apple or any third party ever needing to decrypt your data on their servers.

In theory, assuming we believe what Apple claims, this means your phone or computer runs a model comparing your content to a database of known illegal content hashes or patterns. If the model detects a match, it triggers an alert or sends metadata to Apple or relevant authorities for further action.

Apple’s client-side scanning system runs on certain chips:

This means this kind of scanning capability has been technically possible for years thanks to Apple’s Neural Engine and secure enclave hardware.

While client-side scanning is pitched as a privacy-preserving measure, there are major risks and concerns:

Even if scanning is done locally, the system must have access to all your content (photos, messages, files) in unencrypted form before encryption. This breaks the principle of true end-to-end encryption.

This creates a backdoor, not a traditional backdoor that Apple holds, but a built-in mechanism to scan all personal data on your device, which governments or other actors could exploit.

Machine learning models can make mistakes. Innocent content could be flagged as illegal, leading to unwarranted investigations or censorship.

Apple or third parties could expand the scanning scope beyond CSAM to politically sensitive content, activist messages, or even legal but “undesirable” material.

Governments with authoritarian tendencies could demand Apple or other companies expand scanning databases to include content critical of the government, religious minorities, journalists, or political activists.

Client-side scanning can become a tool for mass surveillance and suppression of dissent under the guise of security.

The scanning algorithms and databases themselves become a target for hackers. If malicious actors access these scanning components, they could 1) Reverse engineer or spoof the scanning process. 2) Introduce malicious hashes to flag innocent content, disrupt user privacy, or frame individuals.

While Apple claims only metadata or alerts are sent if matches occur, client-side scanning creates a system where detailed knowledge about your data exists on your device. This could be leveraged to collect user behavior patterns or metadata, either by Apple or third parties.

Apple or “big tech” could use this data to profile users, refine targeted advertising, or even sell aggregated behavioral data to third parties, despite privacy promises.

Eroding End-to-End Encryption: client-side scanning introduces a “soft” backdoor that can be widened over time, allowing tech companies to access user content under the pretext of safety.

Expanding Surveillance Beyond CSAM: The scanning scope could gradually broaden from child abuse material to drug-related content, political opinions, or copyrighted content, increasing user data exposure.

Selling or Monetizing User Data: Although Apple says it doesn’t sell user data, the existence of detailed metadata or flagged content creates new rich datasets that could be monetized directly or indirectly by advertisers or government contracts.

Creating a Precedent for Other Companies: If Apple successfully normalizes client-side scanning, other big tech players might implement even more invasive scanning or data collection systems, accelerating privacy erosion industry-wide. Government Collaboration and Pressure: Governments can compel companies to add new categories to scanning databases or share flagged data under secret orders, sidestepping traditional privacy safeguards.

While client-side scanning may seem like a privacy-preserving way to combat illegal content, it fundamentally shifts control of your personal data from you to big tech and governments by:

• Forcing devices to scan all private data locally,
• Increasing surveillance capability and risk of abuse,
• Weakening encryption and privacy protections,
• Opening avenues for misuse, false accusations, and government overreach,
• And enabling new forms of data collection, profiling, or monetization.

Here are some of the legal cases from 2025 involving Apple where it has been accused (or found) of violating user rights, privacy, consumer protection, misleading advertising etc.

Keep in mind that these are just the ones I found with a relatively quick search, since compiling a comprehensive list is beyond the scope of this post, and, importantly, this only covers the year 2025, the list would have been too long if I included additional years.

• The Guardian
• Reuters
• Reuters

snippet.cpp

by:
▖     ▘▖▖
▌ ▌▌▛▘▌▙▌
▙▖▙▌▙▖▌ ▌
 
written: October 17 2025