Skip to content

Topic

Platform Trust and Safety

How online platforms and app stores police fraud, scams, impersonation and abuse through review, takedowns, verification and payment protection.

Current stories

securityOne report1 publisher

Shell pages make up 3.64 million of Meta's 3.8 million Singapore scam takedowns

Meta acted on nearly 3.8 million scam-linked accounts and pages in 2026 using Singapore Police Force signals. Dormant shell pages from one July sweep make up 3.64 million of them, so the repeatable result is 20,000 police flags turning into 113,000 takedowns.

Publishers:police.gov.sg

Reality

Evidence40
Adoption50
Hype gap+35
Incentives75
Confidence45
productOne report1 publisher

LinkedIn hands resume checking to your old colleagues and your employer's page admin

The new vouching and brand-control tools give the checking work to colleagues who have known you for a year and to page admins clicking through profiles one at a time. LinkedIn says 115 million members are already verified.

Publishers:techcrunch.com

Reality

Evidence58
Adoption45
Hype gap+25
Incentives72
Confidence60
productOne report1 publisher

404 Media put an AI track on a real band's Spotify page by claiming to be the band

The upload path takes the submitter's word for which artist a release belongs to, so an AI track lands on the real band's catalog page and pays whoever filed it. Spotify's release-review tool has been in limited beta since March.

Publishers:engadget.com

Reality

Evidence66
Adoption38
Hype gap+24
Incentives60
Confidence57
securityOne report1 publisher

Fraudulent copyright strikes suspend Instagram accounts weeks before Meta unwinds them

Meta acts on copyright complaints without checking who filed them, so anyone with an email address can lock an Instagram account that earns money. The BBC found an owner who paid $50 in crypto and was hit again at once.

Publishers:malwarebytes.com

Reality

Evidence55
Adoption30
Hype gap+10
Incentives60
Confidence55
securityOne report1 publisher

Meta's 95% certainty bar turns suspected scam advertisers into premium customers

Documents reviewed by Reuters put Meta's projected 2024 scam and banned-goods ad revenue near $16 billion, while a February 2025 internal plan set aside $135 million in forgone revenue to curb suspicious advertisers.

Publishers:news.risky.biz

Reality

Evidence44
Adoption
Not scored
Hype gap+14
Incentives76
Confidence41
securityOne report1 publisher

A $115 pipeline turns Google's ad transparency archive into a malvertising feed

NYU and Radboud researchers spent a year turning Google's Ads Transparency Center into a detector for deceptive software ads, and found that the report button removes a creative while leaving the domain behind it serving.

Reality

Evidence58
Adoption30
Hype gap+12
Incentives45
Confidence55
securityOne report1 publisher

Two researchers traced a Nigerian sextortion suspect from an Instagram DM to his village

Erin West and Paul Raffile ran a decoy teenage account, logged the IPs behind the sextortion DMs, captured a face through a browser permission prompt, and got to a legal name and GPS coordinates on open tooling.

Publishers:404media.co

Reality

Evidence34
Adoption15
Hype gap+18
Incentives62
Confidence44
buildOne report1 publisher

C2PA proves who signed, not what was true: Content Credentials are provenance, not AI detection

The standard signs a chain of assertions about capture and edits. It does not verify the scene in front of the lens, and a missing manifest tells you almost nothing.

Publishers:dev.to

Reality

Evidence42
Adoption
Not scored
Hype gap−12
Incentives22
Confidence48
scienceOne report1 publisher

One variable beat ideology across six extremist movements in an Australian survey of 3,429

Hostility toward women separated supporters of extremist violence from everyone else across incel, far-right, religious and far-left categories alike, the research team reports.

Publishers:phys.org

Reality

Evidence42
Adoption
Not scored
Hype gap+32
Incentives68
Confidence38