Pinned Tweet
Hey @elonmusk & the @DOGECommittee team, are you seeing dead people get paid? So do we.
And we've solved it.
We recently ran a portfolio scrub to cleanse one of the country’s top 5 banks and found over 200,000 active accounts linked to deceased Social Security numbers. That’s 16 basis points of a 125M account sample, and alongside another 133,000 synthetic identities that were also hiding in plain sight.
At another government agency we cleaned up accounts where they had been paying dead retirees for more than 18 years.
95% of state-reported deaths aren’t in SSA records. So at Socure we uniquely cross reference obituaries, probate, property records, and more to deliver the most accurate (and real-time) record of when an identity becomes deceased + we now verify more than 1 billion identities every month across 2800 banks, fintechs, government agencies, marketplaces and big tech.
Socure can quickly and efficiently help DOGE and the US taxpayers stop the whole paying dead people problem - DM me.
Johnny Ayers retweeted
with all due respect, that is not what agents need.
agents need a way to RELIABLY EXECUTE ON HUMAN INTENT.
i’m bearish on all companies that are giving agents human identifiers. they all seem to think the pain point is that agents are being blocked by all these institutions, websites and merchants, so they make it easier for agents to bypass them.
but we need to think why this is happening!
every one of those blocks exists because a human identifier in the wrong hands is the oldest fraud vector there is:
- the FTC logged $15.9B in reported fraud losses in 2025, up 24% YoY
- identity theft reports hit 1.36M and credit card identity theft alone was up 33%
- javelin puts identity fraud losses at $27.2B
- the FBI flagged account takeover as a rising vector, with $893M in losses already tied to AI
step-up auth, device checks, bot detection. institutions spent decades building these to answer one question: is this really the person who owns this identity?
handing an agent your card number, your login or your SSN doesn't answer that question. it makes the agent indistinguishable from the fraudster.
“I wish I could do what I do with Socure in the U.S.—in every other market I operate in.”
Well, now you can.
Gone are the days of juggling multiple vendors, navigating country-by-country compliance, or drowning your risk team in manual reviews with inferior point solution providers that lack best-in-market coverage and performance.
Socure’s RiskOS™ is now fully global—powering instant, compliant identity, compliance, and risk decisioning across more than 190 countries.
What once took months and multiple partners is now handled in real time by a single platform (RiskOS by Socure) that’s known broadly across the market for only bringing the best-performing solutions to market. Ruthlessly analytical, every outcome measured and optimized.
What’s under the hood:
◼️ eKYC in 190+ countries
◼️ 6,000+ Document ID types, 49 languages + support for bring your own language (BYOL)
◼️ Phone intelligence and phone prefill in 60+ markets
◼️ Real-time KYB, KYC & UBO globally
◼️ Global OFAC, PEP, Adverse Media, Global Sanction & Enforcements Screening & Automated Monitoring across both consumers and businesses
◼️ One no-code orchestration & decisioning platform to write any rule, build any model, write any strategy with an AI copilot
From seed-stage fintechs in LATAM to Fortune 50s scaling across APAC, RiskOS is transforming identity, compliance, authentication and growth at global scale—with results to match:
✅ A neobank raised approval rates from 78% → 96% in the UK
✅ A global marketplace saw 96% phone verification & prefill rates in Brazil
✅ A global employer of records cut onboarding from 4 days → 2 minutes by fully automating their global IDV
One identity graph. Best in market end-to-end IDV, fraud, authentication, and compliance solutions. One platform to trust and serve your users anywhere in the world.
This is how you grow global and FAST.
socure.com/news-and-press/ri…
1
111
Johnny Ayers retweeted
Plaid and Other Aggregators just took a tremendous (well deserved) hit last week: PAY BY BANK IS DEAD IN US. US Banks just set fees for API access/ authentication.
Any "business" model that does not have shared economics will have a short shelf life. Building value on someone else's hard work, creating a competing product to end run them, and keeping all the economics will not gain you any friends. Surprise.. banks are fighting back.
My bank was Yodlee's first customer in 97 (Wachovia), banks can't own all the dross of compliance, customer support, tech, infrastructure, ... with no upside. At one point in time Yodlee represented 30% of our call center volume and about the same in HTTP requests (before we converted them to OFX). There are real costs to be covered. Customer online credentials were never designed to be given away to 3rd parties. Modernization and tokenization of credentials is another cost.
While PaybyBank was already a stretch against Durbin debit rates of $0.21.. The new access fees (vary by bank) will kill this completely. The aggregators may complain, but the banks will have a solid case by demonstrating the real costs involved in supporting this service.
The aggregtors will need to move up market to RTP and FedNow.. but this also means abandoning consumer eCommerce to focus on markets where speed matter (B2B, Payouts).. and competition has a big head start.
The Socure team just released a fascinating report on how fraudsters are using spoofed domain names to impersonate real companies, and how our models are spotting the anomalies to protect them.
Let me start with a true story for reference:
An innocent woman was recently scammed out of thousands of dollars by a fraudster pretending to be a Socure hiring manager. They didn't need a fake website or a fake phone number. All they needed was an email domain name that looked close enough to ours.
@ socure.team
Armed with that email domain, they were able to run a whole ruse convincing her that she was hired at our company and sending her off to do a series of "tests" that resulted in sending significant amounts of her own money to the fraud ring, thinking it was Socure.
It's sick, but luckily there is a pattern.
Fraudsters register look-alike domains in bursts - dozens at a time through fly-by-night registrars, often with zero KYC. They mimic employers, banks, even government agencies. Some spoof emails, some fake job offers. All are designed to disappear before defenses kick in.
What stands out?
Volatility.
Legit domains (like gmail) behave consistently (see image 1)
Fraud domains spike, vanish, and reappear under new names (often within days). That chaos provides a clear signal and Socure's models are built to spot it (see image 2).
Our AI-driven RiskOS platform is able to identify the patterns across domains, devices, and applications. If you look at the charts, they luckily stick out like a sore thumb, allowing us to alert customers of the potential fraud before it occurs.
Full blog post: socure.com/blog/shape-shifti…
2
149
Top Market Intelligence Firm, Liminal, just featured Socure as industry leader in Global KYB/KYC - now available in 190 countries.
The pitch is simple: Global companies want a single platform that can manage their KYC/KYB in any country in the world. We have been investing massively in our coverage, partnerships, and feature set for the RiskOS platform to be that single go-to source.
“Socure’s ranking in our recent report underscores its accelerated leadership in KYB innovation. We were particularly impressed with Socure’s ability to provide an integrated BEV solution combining business verification, identity validation, and fraud prevention within a single API.” - Will Charnley, Chief Operating Officer of Liminal.
1
70
We just had another piece of Socure research published in Fortune -
This time its detailing how scammers stole $13 million in student grants last year.
In one of our recent tests with a major state college system, we found that as many as 1 in 5 applications currently being submitted are fraudulent.
The scam?
Fraudsters apply for admission using stolen or synthetic identities. Then, they apply for government grants (with the acceptance letter as proof of “legitimacy”). As soon as the grants hit their account...
Buhbye.
Given that historically this was not a fraud vector that colleges systems were aware of, they largely lack an effective process for screening out fake applicants. Exacerbate that with a student loan budget standing at $3.2B, that is a really large honeypot that scammers are now fully aware of and increasingly attacking.
It's not all bad news:
Our models were able to detect 98% percent of the fraudulent applications. Many are sloppy with dead giveaways like gibberish email addresses, bogus IPs, overseas geolocations, originating from device farms and invalid phone numbers. Meanwhile, there were many other fraud patterns that we stopped that were much more complex.
Full story in Fortune: fortune.com/2025/06/12/ghost…
Scammers are incredibly innovative. We’ve had fake candidates show up with AI-written resumes, altered work history dates, synthetic identities, sophisticated deepfakes and LinkedIn profiles that disappear mid-process.
One guy made it all the way to a live interview. He was a smooth talker with polished answers, great resume, until we realized he was reading straight from ChatGPT during a request to sync Excalidraw. We could see his eyes reading, the prompt > answer matched the exact responses of 4o.
Nice try. He also had an exact replica of the background that Chamath has from All-In.
We run every applicant through the same systems we use to stop identity fraud in banking, marketplaces, gaming and government. Email and phone verification. Synthetic and ID fraud models. Device fingerprinting, geolocation and behavioral analytics. 1000s of real-time passive signals that catch fake and stolen IDs before they ever reach an interview. We also have a hosted UX flow for companies to drop in wherever they'd like in their flow for physical identity verification, liveness verification in a fully-automated, 2 second process.
Here’s what we’ve learned:
◼️ Real candidates have email histories that go back years. The fakes, typically, a few weeks.
◼️ Real candidates don't have telemetry between sessions of 100s of miles apart when switching from mobile to web.
◼️ Real candidates don't mismatch the name. They're not on VPNs, emulators, higher risk networks and they rarely have some form of technical difficulties.
◼️ Real candidates generally don't vanish the second we ask for additional information or physical document verification to be verified in session.
U.S. intelligence says North Korean operatives have already infiltrated companies across tech, finance, and defense, funneling 100s of millions back into sanctioned regimes. They’re targeting remote roles with access to sensitive codebases, consumer data, and IP.
It’s a strategic infiltration by nation-state actors, this is a national security risk.
Most hiring systems and recruiters weren’t built to verify identity ownership, and certainly not with the level of sophisticated AI attacks we're seeing with deepfakes and injection attacks. I-9 checks documents. E-Verify checks format. None of these confirm the person is real, and that it is the correct and same person in each session, over time.
We’re sharing what we’ve learned so others can tighten their defenses. If you’re hiring remotely and skipping identity verification, you’re leaving the door wide open. We'd love to help ensure your company doesn't become a Wall Street Journal headline.
1
64
We just published a report that sounds the alarm on the scale of fraud attacks on our government sector programs. As much as 12% of applications for federal funding and loans are believed to originate from foreign fraud rings.
During the pandemic, it was estimated that we lost over $500B in PPP payments to foreign adversaries.
If you thought that was a one-time thing, you’re sadly mistaken.
That was the canary in the coal mine.
This report goes into a tremendous amount of detail as we dove deep into the data around fraud ring activities targeting US government programs.
Just a few highlights:
◼️ The international fraud groups are originating from China, Russia, Poland, India, South Africa, Philippines, and several other nations.
◼️ Fraudsters target multiple agencies at once.
◼️ Identity theft is far more prevalent than synthetic IDs.
◼️ Fraudsters attack government and commercial entities with the same stolen identities.
There's an incredible amount more detail inside the report. I highly recommend a read: socure.com/news-and-press/so…
Once you understand the patterns, it will become far easier to know what you need to be prepared for.
And if you need help, of course, please reach out.
1
40
A Real Estate Agent. Nearly 200 applications. One Detroit address.
What do they all have in common?
Christmas Eve birthdays and a sprawling synthetic fraud operation that’s still active.
Our team uncovered a fraud ring centered around one Detroit household, run by a family that’s quietly built and maintained more than 30 synthetic identities since 2021.
The signals were subtle, until they weren’t:
◼️ A surge in December 24 birthdates across fabricated and manipulated identities
◼️ Reused Gmail patterns with randomized numbers
◼️ SSNs tied to the deceased
◼️ A rotation between prepaid and VoIP numbers to slip through checks
◼️ IP addresses jumping from Michigan to Texas and California
We’re calling this case The Detroit Family Synthetics - and it's a prime example of how these fraud rings evolve. They get smarter, learn from what works, and change tactics fast.
1
62
At Socure, we talk a lot about growth, but few people embody that word like Rivka Gewirtz Little.
She joined us as a bold, strategic thinker and quickly became the connective tissue between executive vision and company-wide execution. As my Chief of Staff, she’s helped drive our go-to-market strategy, ensured cross-functional alignment on innumerable strategic initiatives (actually so many its incredible), and honestly just been the best partner to me I could have ever imagined. As a CEO advisor she was calm (mostly), wise (beyond her years) and as my Chief of Staff, she most critically had my back (all the time).
So it is with great pleasure to announce to the world that Rivka has been promoted to be our new Chief Growth Officer a role that not only reflects what she’s already been doing but expands her impact even deeper as we scale into the next chapter of 40%+ revenue hypergrowth (with goals to exceed 50).
However, her journey is a masterclass in how growth isn’t just about numbers. It’s about People. Strategy. Vision. Leadership. And most importantly, a desire to create the future of identity.
She’ll lead a newly formed Growth & Insights organization where market strategy, fraud intelligence, storytelling, and public policy converge to shape the next phase of Socure’s history.
There are few leaders who can both see the big picture and move the operational 3D chessboard like Rivka.
1
60
Just launched a new episode on the Fintech Leaders Podcast with Miguel Armaza. Check it out.
We discussed:
☑️ How Socure rebuilt in 2014 and turned a near disaster into an advantage
☑️ Why government fraud prevention lags 15-20 years behind the private sector, costing billions to tax payers
☑️ Our strategic approach to M&A, including their acquisition of Effectiv
☑️ Preventing $3 billion in fraudulent disaster relief applications in just six weeks
☑️ The future of digital identity: from biometrics to tokenized credentials and wearables... and a lot more!
Link in comments...
66
How do you transform a company with hundreds of millions in ARR, profitable, trusted by thousands of customers across 40+ industries... to be AI-native without breaking (too much)?
That’s the question we’ve been asking ourselves at Socure.
As the world has transitioned relatively slowly over the last 20 years from software-native to internet-native to mobile/cloud-native to Data/ML-native to what is now an extremely accelerated transition to AI-native, it requires an equally rapid rethinking of every core function.
AI-native means AI is core user interface, engine and product differentiator of the company vs just a feature bolt-on.
In order to do this, every team, function, product, and capability must be rethought and redeveloped from the ground up across:
◼️ Engineering, data science, product
◼️ Storytelling, content, design
◼️ RevOps, finance, legal
◼️ Sales, customer success, support
We’ve been building machine learning models for over a decade. We’re deeply technical. We’re analytical at our core.
But the bar has moved with AI and very, very fast.
What used to be world-class SaaS at $300K ARR per FTE is now world-class AI-native at millions of $s per FTE because of AI agents, co-pilots, autonomous workflows and with near zero marginal cost.
From the outside, Socure is already the market leader. We’re growing quickly, profitably, and hands-down, objectively, have the best performing solutions in our market.
Conventional wisdom would say: if it ain’t broke, don’t fix it.
But we’ve never been conventional.
To create a generational company in identity, compliance, credit, and fraud prevention, we have to constantly reinvent ourselves. The speed and magnitude of change with AI is different.
So we've quadrupled down and launched a company-wide initiative called "AI Everywhere" which means:
👉🏼 Full alignment from Board and Leadership Team to go all-in on AI
👉🏼 Asked every team member to get 20% more productive in less than 6 months
👉🏼 Planned multiple internal hackathons across every org with $25K grand prizes
👉🏼 Identified 8 strategic initiatives where AI could create step-function gains with the largest company-wide impact
👉🏼 Encouraged every Socurean to experiment, build, and fail fast
It’s a full mindset shift.
It means every person in the company becoming the CEO of their own domain, with the full power of the world’s information and tools at their fingertips.
It changes how our organization thinks, operates, and solves problems.
One percent better, every day.
1
2
80
Identity fraud prevention at an inflection point: complacency is no longer an option.
biometricupdate.com/202503/s…
39
Johnny Ayers retweeted
Two words: FedRAMP® Authorized.
🦅🇺🇸💪
Here about it from our Head of Public Sector, @JordansRules.
Fraudsters are about to move from stealing unemployment checks to being on the unemployment line. Socure is now officially FedRAMP Authorized.
That means the same identity verification and fraud prevention platform that’s already powering over 30 state agencies and stopping fraud at record-breaking rates is now certified secure enough for the federal government.
This was a two-year marathon.
Millions invested.
Every function involved.
Because FedRAMP is about proving you can handle the weight of the most sensitive data on the planet.
This was monumental effort across the entire company. Proud of this team! It's a big day for Socure!
A person makes a purchase on a credit card with the premeditated intention of saying it wasn’t them, who eats the loss?
The burden of proof is on the bank or the business, and in the vast majority of cases the fraudster gets away with it. It's called first-party fraud, and it's a $100B a year problem.
Trying to prove first-party fraud after the fact is a poor solution, too manual. Often times, it's cheaper to eat the fraud than to fight it, and very few firms have the operational capacity to handle the # of disputes they get each day, especially when under attack.
The key to slowing the financial losses and operational burden is being able to stop repeat offenders before the transaction occurs.
Historically, banks and fintechs were unable to do this on their own because the fraudsters' prior incidents will have happened on some other bank's card. We regularly see banks hit by first party fraudsters that have done the same thing across 10, 15 or even 20 other institutions before them. But because banks and fintechs do not share data broadly, in many cases the bank's experience with this fraudster is a first-time incident.
This is where Socure can help organizations get an edge. We validated over a billion identities in December, across 2,800 institutions, including banks, fintechs, government entities, e-commerce companies, social media platforms, and more.
By tying a unique identifier called the Socure ID to every identity across our graph, we're able to recognize first-party fraud patterns, tie them to the SocureID and warn the next institution that a fraudster is trying to hit them is a high risk.
That’s the story behind our new press release, announcing the first of its kind: The Identity Manipulation Risk Score - specifically targeting repeat first party abusers who are trying to hide or alter their real identity to steal funds.
The power of network efforts, realized in partnership with Socure. Ping me if you want to learn more.
Press release in comments.
50
Today, we’re launching Socure's RiskOS™ - The First AI-Native Decisioning Engine Built on the Industry’s Most Comprehensive Identity Graph.
As sophisticated fraud attacks continue to cost businesses hundreds of billions of dollars each year, fragmented strategies and isolated point solutions are no longer enough. RiskOS™ delivers the industry’s most powerful AI platform, capable processing tens of thousands of real-time computations in milliseconds and launching with over 50 pre-integrated third-party data solutions.
Streamlining risk, fraud, and compliance operations through its no-code interface, making strategy changes or adding new third party solutions to any workflow, at any point across the customer lifecycle, can be done in just minutes vs weeks or months.
Here’s the breakdown -
61
2024 will be the year I remember where our entire team really started to fire on all cylinders as a single unit.
“Individually, we are one drop. Together, we are an ocean.” Ryunosuke Satoro
We just released the headline numbers illustrating what this team has accomplished & I'm so excited to share with you.
OUR GROWTH:
▪️ 2.7 Billion Identity Requests Verified (370M unique identities, doubling 2023 totals). We are currently verifying every US citizen ~3x per year.
▪️ Customer base grew 43% YoY to over 2,800 organizations across financial services, fintech, state, local and federal government, e-commerce, insurance, big tech, crypto, AI & more, spanning ~35 different consumer verticals.
▪️ 193% public sector growth, now serving 13 states, 30+ state agencies, 2 federal agencies & 20+ higher education institutions.
▪️ Closed the year with 54% GAAP revenue growth YoY, surpassing all goals while delivering attainment of 108% of new bookings & 121% adjusted operating profit vs plan.
OUR ML MODELS PERFORMANCE:
▪️ Sigma Fraud Suite: Caught more than 92% of ID fraud in just the riskiest 3% of all users (industry avg: 37%) while reducing false positives by more than 40% YoY between Sigma ID v3 and v4.
▪️ eKYC/Identity Verification: Verified 99% of all US consumers, with a 3% YoY improvement, including 30% more 18-year-olds and 92% of Gen Z compared to any other competitor. Internationally, we now serve customers in more than 70 countries with a minimum of 80% coverage, with many countries well above 90%.
▪️ DocV: Industry-leading 98.2% true accept rates with market-leading verification speeds of 1.5 seconds (p99), with only basis points of false accepts. Secondary documents launched meanwhile selfie reverification continues to be deployed broadly across our DocV user base for ongoing authentication.
The list is too long! Socure ML model outperformance continued in email & phone coverage and fraud capture, sanction screening name degradation accuracy, device to SocureID knowledge graph pairs, bank account verification coverage and accuracy, first-party fraud consortium contributions, identity manipulation scoring & more.
OUR INNOVATIONS:
▪️ Filed 13 new patents (two granted for fraud prevention and document verification).
▪️ Launched advanced prefill capabilities, achieving 90%+ first-time rates, outperforming competitors by more 30%.
▪️ Reduced sanctions screening times from 15 minutes to less than 1 minute using our GenAI explainability agent for Compliance.
OUR AWARDS & RECOGNITIONS:
▪️ Leader in the 2024 Gartner® Magic Quadrant™ for Identity Verification.
▪️ Winner of the Finovate Best Generative AI Solution for Sigma v4.
▪️ Fraud & AML Impact Award for Identity Verification Innovation by Datos Insights.
Looking back, I’m beyond grateful to our customers, partners, & the incredible Socure team for proving that performance, speed & inclusivity can go hand in hand.
And we have a BIG launch next week! I can't wait to see what this team can achieve in 2025.
Link: socure.com/news-and-press/so…
53














