HN Debrief

Never Give Them Your Face

  • Privacy
  • Regulation
  • Security
  • AI
  • Identity

The post is a broad warning against giving websites and platforms your face for “age verification.” Its core claim is simple: most of these systems are not merely checking whether someone is over 18. They are collecting identity documents, face scans, and persistent personal data that can later be repurposed for tracking, policing, or political control. The piece pushes a hard line of refusal, from porn sites to airports, on the theory that once biometric verification becomes normalized for low-stakes online access, it becomes easier to require it everywhere.

Treat biometric checks as a product and policy choice, not an inevitability. If your company touches onboarding, moderation, payments, or identity, assume users increasingly care about revocability, appeal paths, and whether you can solve the problem without collecting immutable data.

Discussion mood

Mostly supportive of the anti-biometric argument, but frustrated and fatalistic. People were angry at broken platform verification flows and mission creep, while also feeling that face-based identity checks are already embedded deeply enough that individual refusal only works at the margins.

Key insights

  1. 01

    Biometrics make account recovery impossible

    Using a face as a credential turns ordinary trust-and-safety bugs into dead ends. The Facebook signup stories show the failure mode clearly. A user can hand over a face scan, still get locked out, and then lose the one thing that usually makes broken systems survivable, the ability to start over with a new account and rebuild trust.

    If you design identity systems, keep a revocable recovery path that does not depend on the same biometric you already rejected. For users, avoid services that combine irreversible identity binding with no meaningful appeals process.

      Attribution:
    • __MatrixMan__ #1 #2
    • howard941 #1
  2. 02

    Reputation systems are banning infrastructure, not people

    A lot of these lockouts are not even about the face scan itself. They are caused by crude risk scoring on VPNs, IP ranges, device fingerprints, or opaque policy flags. That means users get pushed into stronger identity checks only because upstream heuristics are noisy and platforms want the cheapest way to reduce false negatives.

    Watch for teams using biometric verification as a patch over broken abuse detection. Improving attribution and appeals may reduce pressure to collect stronger identity data in the first place.

      Attribution:
    • jazzyjackson #1
    • a2128 #1
    • ranger_danger #1
  3. 03

    Privacy tools only work when normal people use them

    The Tor discussion landed on a blunt network effect point. An anonymity system protects users better when it is common enough that using it is unremarkable. The Harvard bomb-hoax example was cited as a case where unusual Tor usage itself became a clue, not a technical break of Tor. That same logic was applied to iCloud Private Relay, which helps by making IP obfuscation mainstream rather than niche.

    If you want privacy-preserving infrastructure to matter, optimize for default or mass adoption, not just technical purity. Tools that only experts use can become self-identifying.

      Attribution:
    • judge2020 #1
    • jupr #1
    • bronlund #1
    • john_strinlai #1
  4. 04

    The policy demand is real even without a villain

    Several comments cut through the all-purpose conspiracy framing. Parents, regulators, and many ordinary voters genuinely want stronger restrictions on kids' access to parts of the internet. The hard problem is that once you demand high-assurance online age gating, identity verification keeps sneaking back in because the weaker alternatives are easy to share, spoof, or route around.

    Do not build your strategy on exposing bad motives alone. If you oppose mandatory age checks, you need a credible answer to the underlying child-safety demand, even if that answer is narrower device controls and better parental tools.

      Attribution:
    • Aurornis #1
    • andrewla #1 #2
  5. 05

    The leverage point is regulation and appeals

    The practical fixes people kept reaching for were legal, not technical. The EU Digital Services Act was cited for requiring complaint and redress mechanisms. Others argued that the bigger gap is modern privacy law and limits on data brokers, plus rules forcing human contact and appeals when automated systems cut someone off from major platforms or markets.

    If your business relies on automated identity or trust decisions, expect regulation to focus on process rights as much as data collection. Build explainability, escalation, and human review before they become mandatory.

      Attribution:
    • lenerdenator #1
    • flipbrad #1
    • jkestner #1
  6. 06

    Small business access is collateral damage

    A recurring complaint was that Meta's verification and account integrity machinery is now bad enough to block legitimate businesses from creating pages, running ads, or even opening fresh accounts. The notable point is not just privacy harm. It is that identity friction is now directly suppressing commerce for people trying to give Meta money.

    For platform operators, abusive-user controls that choke honest new entrants are a growth problem, not just a trust-and-safety problem. For founders, dependence on a single gatekeeper platform remains an operational risk.

      Attribution:
    • chamomeal #1
    • ferngodfather #1
    • maaarghk #1
    • dd8601fn #1

Against the grain

  1. 01

    Some identity checks are legitimate

    Not every face or ID request belongs in the same bucket. Banks, brokers, governments, and border systems often do need strong identity because they are authorizing money movement, legal status, or state action. The sharper distinction is between high-stakes institutions with clear need and random websites adopting the same machinery for convenience or data capture.

    Be precise when arguing against biometric collection. You will be more credible if you separate low-stakes age gating from contexts where strong identity is genuinely part of the service.

      Attribution:
    • andai #1
    • nickelpro #1
  2. 02

    Photo proof does not itself prove bad faith

    The strongest pushback on the article's logic was that a bare document saying 'over 18' is useless unless the verifier can tell the holder is its rightful owner. In physical spaces, that is exactly why photo ID became the norm. The digital privacy risks are much worse, but that does not by itself prove that every request for a face match is secretly about surveillance rather than age.

    When evaluating a verification flow, ask what assurance level the operator actually needs and whether the retained data matches that need. Bad policy often starts with a real authentication problem and then over-collects.

      Attribution:
    • hennell #1
  3. 03

    For many people the privacy breach already happened

    A fatalist line ran through the comments that face collection is already everywhere. Phones, passports, airports, retail cameras, social networks, carriers, and mortgage systems already tie identity, location, and images together at scale. From that view, refusing one more facial check does little unless paired with broader limits on data brokerage, retention, and cross-system linkage.

    If you work on privacy policy or product design, focus not just on collection events but on downstream use, retention, and interoperability. Stopping one scan matters less if every database can still be joined later.

      Attribution:
    • SoftTalker #1
    • marcta #1
    • w4yai #1
    • schrodinger #1
  4. 04

    Pure refusal is not a complete answer

    Some readers thought the article loses force because it tells people to resist without offering a viable replacement for protecting children online. They argued that waving away the harms of social media or dumping the whole issue on parents will not persuade lawmakers who already see a real social problem. Opposition needs a concrete alternative, not just a veto.

    If you are lobbying against age-verification mandates, pair the objection with implementable substitutes. Device controls, browser ratings, parental tooling, and limited-scope age signals are more persuasive than blanket rejection.

      Attribution:
    • kspacewalk2 #1
    • sailfast #1

In plain english

digital credentials
Cryptographically signed digital documents that can prove facts such as age or identity online.
Digital Services Act
A European Union law that sets rules for online platforms, including moderation, complaints, and user protections.
EU
European Union, the political and economic bloc of European countries.
iCloud Private Relay
An Apple service that hides a user's IP address from websites by routing traffic through two separate internet relays.
IP
Internet Protocol address, a network identifier often used to locate or classify a device on the internet.
Meta
The company that owns Facebook, Instagram, WhatsApp, and other products.
Tor
A privacy network that routes internet traffic through multiple volunteer-run servers to make it harder to trace a user's location or identity.

Reference links

Policy and regulation

Background on lobbying and coordination claims

Privacy and verification tech

Cases and examples