Kalalau Trail, 2015 - Rob Leathern
Hello, folks. Jonathan here. Rob Leathern and I worked together at the end of my Google career, when he led Privacy product management. He writes regularly about similar issues as Platformocracy, from an abuse and transparency lens. He has also been tracking the Meta ad fraud leaks, so we’ve traded newsletters for the week. You can see my piece this week, on why platforms consistently underinvest in safety, over at Rob’s Notes.
We are living in the Golden Age of Scams. Maybe I knew it when I saw the police video of a woman who’d fed over $20,000 in $100 bills into a bitcoin ATM machine. The officer stopped her from sending her savings as crypto to scammers, though many others haven’t been as lucky, despite companies like Coinstar helpfullying saying they’re “committed to consumer protection”.
I started a nonprofit called CollectiveMetrics.org which was featured in Wired last week. We believe that the best way to improve accountability for the large platforms that run our lives is to lean on transparency and create open-source measurement systems and public trust metrics.
I spoke to a nonprofit leader recently whose group has done some amazing work dissecting scam ads using the public ad libraries Meta and Google provide in the EU and elsewhere. Unfortunately, they think these companies are in fact going to withdraw the transparency they currently provide into ads outside of the EU, because in general (with some exceptions) national laws do not in fact mandate this level of transparency into ads. That would be a horrible backwards step, but in the absence of meaningful regulation in the US, I can also see how they got there.
Fraud-specific regulatory interventions in the US have followed a pattern over the last 125 years, dating back to the lottery scandal of the 1890s, patent medicines in the early 1900s that led to the creation of the FDA, Ponzi's scheme (1920) that strengthened state blue sky security laws, the 1929 market crash that led to the SEC, the S&L crisis (1980s) etc. In short:
Phase 1: Innovation & Exploitation
Early operators discover profitable loopholes. Mix of legitimate actors and bad actors. Regulators view it as "market innovation."
Phase 2: Industrial-Scale Fraud
Bad actors optimize and scale. Playbooks spread. Legitimate players exit or get corrupted. Vulnerable populations systematically targeted.
Phase 3: Regulatory Paralysis
Regulators have tools but don't act. Industry capture, jurisdictional confusion, or "it's too complicated" excuses. Occasional enforcement actions that don't dent the problem.
Phase 4: Crystallizing Crisis
A scandal too big to ignore - suicide victims, wiped-out retirees, congressional hearings. Media coverage makes inaction politically impossible. Cries of “how did we let this happen?"
Phase 5: Bright-Line Overcorrection
Sweeping bans, strict liability, personal criminal penalties for executives. Industry protests "this will destroy innovation." Usually it doesn't.
I’d argue we're in the late "regulators have tools but lack will" phase, heading toward a crystallizing event that triggers the "bright-line ban" phase. The nonprofit we started is either (a) our attempt, with plenty of inside knowledge, to move towards building tools that help industry self-regulate itself better and perhaps avoid bright line bans… by opening content to third parties (like us) who can measure how good of a job companies do, and/or (b) the data we find will help make it clear that companies just simply are NOT doing enough.
These companies and their shills will make the spurious argument that consumers have agency - while the same firms do nothing to protect the vulnerable. The aforementioned historical fraud schemes systematically identified and exploited vulnerability markers:
Patent medicines ("snake oil") targeted immigrant communities with foreign-language newspapers, women who were excluded from medical establishments, and rural populations without access to doctors
Penny stock boiler rooms in the 1990s like Stratton Oakmont (think “Wolf of Wall Street”) built "sucker lists" (literally what they called them) of retirees with nest eggs, recent widows, and people who'd responded to previous investment mailings
The S&L crisis in the 1980s saw operators like Charles Keating specifically target retirees through hotel seminars, church groups, and "senior appreciation dinners." Lincoln Savings sold $250 million in worthless bonds to 23,000 mostly elderly customers who thought they were buying insured CDs
Lottery frauds in the 1890s specifically advertised in Southern Black churches and newspapers, exploiting post-Reconstruction poverty
Powered by state of the art machine learning and AI, scam ads do this kind of vulnerability exploitation today at greater scale with far greater algorithmic precision.
By itself, I’m not sure that the Reuters story about 10% of Meta’s ads being fraudulent will end up being that crystallizing crisis event. While I was surprised an internal number like that would be that high, I also don’t know what’s in that figure and not in it. Ideally nonprofits like CollectiveMetrics could come up with an independent number like that for every platform every month… so we will continue to fight for access to be able to do just that. Regardless of what any of us write or say, we all still have a lot of work to do to build real solutions that understand and then reduce scams and fraud. I plan to work on this via both nonprofits like CM (if you want to help, please donate or better yet help us build things: e-mail me robl (at) collectivemetrics.org!), and via for-profit businesses we incubate at Hawkview Labs.
I’m sure Jeff Horwitz has more stories to write about Meta. Other journalists and former insiders like Jonathan here are doing great work to shine a light on the downsides of everyone being connected to everyone else.
Because maybe an “overcorrection” is long overdue.
Ideas? Feedback? Criticism? I want to hear it, because I am sure that I am going to get a lot of things wrong along the way. I will share what I learn with the community as we go. Reach out any time at [email protected].
