A think tank says age checks make kids less safe, and the data was buried

A think tank's report became an article became an HN thread

The headline that surfaced in the thread — age verification tech could put children at greater risk — is a news story about a report. A commenter dug up the actual FIPR response document so readers could skip the telephone game.

What the FIPR report actually argues

The report comes from the Foundation for Information Policy Research (FIPR), with input from academics including Alice Hutchings (Cambridge) and Steven Murdoch (UCL). Its core finding is that mandatory age verification both fails its stated goal and manufactures new harm:

• It creates fresh attack surface. To verify age, users hand over biometrics, credit cards, or government ID — pools of sensitive data that enable blackmail and identity theft. FIPR points to Facebook having reused phone numbers collected for verification to target ads.
• It excludes people. Facial-age estimation performs worst on minority, disabled, and LGBTQ+ users, who get locked out of ordinary services.
• It is trivially bypassed. Children route around it with VPNs, deepfakes, and fake credentials FIPR says sell for as little as $0.80 — and end up on less-regulated sites with worse content than the compliant platforms they were pushed off.

FIPR's alternative is 'tagging and blocking': sites self-label content and audience suitability, film-rating style, and parents set the threshold on the device — no central surveillance pool required. Chair Ben Collier's framing: 'while it is tempting to rely on "magic" technological fixes for online harm, these will not work.'

The buried number: the '9 in 10 parents' stat doesn't survive contact

The most valuable reversal in the thread came from a commenter who traced the government's own '9 in 10 parents support a ban' line back to source:

"89% supported "a legal requirement … to have a minimum age of access". However, what the government (and the media) are NOT reporting is that the consultation also paid an independent firm to undertake a nationally representative survey of adults in the general population." — an associate professor of medical physics

The headline figure came from self-selected consultation respondents; the representative survey commissioned alongside it told a different story and went unquoted. A commenter familiar with the policy world also vouched for the messenger, heading off the reflexive 'think tank = astroturf' read:

"FIPR themselves are fine: one of the academic:policy bridge groups rather than an astroturfing organisation funneling dark money … good faith actors." — a commenter who works near UK tech policy

The thread's hard disagreement: is tagging even enforceable?

FIPR's tagging proposal got a frosty technical review. A very-high-karma engineer argued voluntary labels are a fantasy at scale:

"Those can only work if there is some way to ensure that everyone uses them correctly. I guarantee there will be many sites that don't - every single week the kids will discover a new one and spread it to all their friends …" — a long-tenured engineer

A counter-camp wanted the liability moved entirely off parents and onto platforms — make it illegal to advertise to minors, since the ad networks already infer who is a minor, and fine the platforms the way broadcast TV was fined for showing nudity. And several parents described the real bind that makes a ban politically irresistible: a single forbidden child becomes a pariah, so it is a collective-action problem, not a parenting-failure problem.

"once you give a teen a smartphone your control over what websites they visit ends … cutting one kid off from social networks ostracises them. The parents recognize it's a collective action problem." — pjc50, a long-standing UK commenter

The uncomfortable conclusion

The thread and the report converge on the same trap: the measure most likely to pass is the one most likely to make things worse — a centralized ID-and-face pool that kids hop over in an afternoon while everyone else's biometrics sit waiting to leak. FIPR's tagging idea is cleaner in principle and probably unenforceable in practice. The honest read is that there is no clean technical fix here, and the loudest political case for one rests on a survey number that, examined closely, doesn't hold.