
Dreamer of ice cream, cookies, and candy --- Head of World ID @tfh_technology @worldnetwork --- McGill | Northeastern | MIT | Xoogler
Mountain View, CA
Joined November 2010
- Tweets2.1K
- Following1.8K
- Followers468
- Likes9.9K
AI is the least capable it's ever going to be today. Proving that we are human online is the new table stakes to keep the internet useful for all of us. Sandro's post and whitepaper / blog really goes deep on why the world needs it, and why our approach is the future.
1/ Proof of Human is becoming increasingly critical. In the limit democracy and human agency depend on it.
But building Proof of Human is unexpectedly challenging.
FaceID doesn't prevent one person to fabricate human presence for thousands of AI agents.
Government ID based Proof of Human is a surveillance risk and only 1b out of 8 billion people have verifiable IDs.
An anti-surveillance and effective Proof of Human that actually empowers people requires new technology [paper linked in thread] 🧵
Enterprise security can’t stop at verifying the human at the front door.
A sensitive action may cross many systems, each with its own access controls. A valid credential at one hop shouldn’t be enough.
We need a cryptographically verifiable chain back to the authorized human, with policy enforced at every layer.
This is critical for enterprise security, and especially important as agents act across the internet on behalf of humans. We need to know who authorized an agent and what it was authorized to do.
4
132
More people in the US should be paying attention to Japan’s cyberattacks.
Nitori says up to 37.6 million personal records may have been exposed. The risk doesn’t end at the breach: that information can make impersonation and fraud more convincing.
This isn’t just Japan’s problem.
tdnet-pdf.kabutan.jp/2026110…
4
115
Ajay retweeted
What we shipped this week @google
Monday: Two new 8th-gen TPUs
Tuesday: Gemini 4 Carbon
- AlphaProof Code verified Signal protocol
- Gemini Robotics 2.5
Wednesday: AlphaGenome 2 for disease research
- Gemini Decisions API in AI Studio
Thursday: Gemini 4 Hydrogen
- intelligent eyewear pre-orders
Friday: U2 new album (ok not us but exciting!)
onwards!
Ajay retweeted
“It takes 20 years to build a reputation and 5 minutes to ruin it.”
Buffett tried to tell them all and no one listened
Ajay retweeted
“My prediction is that in the next three months, it will be completely impossible to spot that you’re talking to a deepfake.” - @alexblania, co-founder, World
@world_id helps build trust by preventing deepfakes at the source.
As deepfakes become more sophisticated, a familiar voice or face on a screen is no longer enough to establish trust. Here’s one recent case where deepfake technology was used to steal $129M.
Trust infrastructure is now a necessity. World ID for Zoom helps verify that the person on screen is a real human, not a deepfake.
Following an investigation by FBI New York and @SDNYnews, Jacob Soall, Sharon Hazan, and Guy Sheetrit, were charged for their participation in a scheme to use super intelligence (SI) to defraud companies of more than $100 million. Jacob Soall, Sharon Hazan, and Guy Sheetrit allegedly used SI to impersonate the voices and/or images of corporate executives at numerous companies and attorneys at major law firms in order to trick employees into sending large payments to bank accounts all over the world controlled by members of the conspiracy.
If you believe you may be a victim of the conduct described in the press release linked below or know of information regarding its perpetrators, please use the online tips page at tips.fbi.gov.
Read more at: justice.gov/usao-sdny/pr/thr…
Ajay retweeted
.@world_id is trending in Japan:
Nikkei Trendy's "2027 Hit Predictions" listed Proof of Human as the #1 pick
newsonjapan.com/article/1512…
Ajay retweeted
i'm fascinated by human-only spaces inside ai interfaces.
so i built an arcade where you can play together with friends while your agent work.
try it in claude or chatgpt
AI is enabling even more sophisticated hacks into financial institutions
- one solution is employ ‘proof of human’ in the loop
- to prevent agents/ password infiltration
Not a guarantee to mitigate
$WLD @worldnetwork is the only proven proof of human platform
$ORBS @Iamhuman_ORBS
🚨 BREAKING: Hackers are hitting banks in 7 countries at the same time.
This is what Covid looked like in week 1. A few countries. Small numbers. Nobody called it a crisis.
1/ South Korea: hackers used a Chinese AI tool to break into at least 7 financial firms. 68,000 people's data stolen. 16 firms breached by mid-October. This week Daishin Securities reported 38,490 customers' data leaked, bank account numbers included.
2/ Latin America: one gang ran AI agents built on Claude, Grok and Kimi against 400+ targets, banks and payment firms included. They stole passwords and got into customer bank accounts.
3/ Italy: a fake WhatsApp from "the CEO" and an AI-cloned voice got Intesa Sanpaolo's private bank to send out €95 million. €36 million is still MISSING, turned into crypto.
4/ India: small lenders lost ₹300 crore (about $31 million) in 2 months. 150 cases. Hackers copied bank logins and sent the money out as fake loans. One firm was hit twice.
5/ Crypto: $351 million drained from Bitget's hot wallets. North Korean hackers are suspected.
6/ Japan: hackers hit GMO Internet's currency trading service and demanded $50 million in crypto. The prime minister promised an emergency anti-cyberattack package.
7/ Australia: the regulator warns AI attacks put more than A$4.5 trillion in retirement savings in the firing line. Government cyber alerts jumped 80% in a year.
Now zoom out.
AI hacking tools are now easy to get. Smaller firms have weak security. The money moves in minutes. It jumps across banks, through rented accounts, out as cash or crypto. The bosses are often abroad.
Covid locked the streets. Nobody is calling this a crisis yet. But the central banks already plan for the day the hackers get in.
The G7 agreed to run drills that simulate AI attacks on banks. The ECB reportedly gave eurozone banks 4 months to hand in their defence plans.
If your savings sit in a bank, hit follow. I'm tracking where the hackers strike next and how much they take.
Readers added context they thought people might want to know
The listed incidents occurred at different times over months (e.g. Italy in Feb 2026; Bitget Sep 24; Latin America targets May-Sep), not simultaneously as presented.
tomshw.it/altro/95-milio…
decrypt.co/379350/bitget-…
aichatdaily.com/ai-security/ha…
m.economictimes.com/et-front/promp…
smh.com.au/politics/feder…
theverge.com/ai-artificial-…
en.yna.co.kr/view/AEN202610…
Advertisers pay to reach people, not bots.
Proud to join Hakuhodo DY and LG Electronics in Tokyo 🇯🇵 today to build the Human-Verified Ad Network with World ID.
Greater trust in advertising. Better value for businesses. Privacy for people.
Starting in Japan, with global ambition.
本日、博報堂Dホールディングスにて「人間認証型広告」事業を展開するAds for HumanityとLGエレクトロニクスと共に、AI時代の“人間のための広告”へ、「人間認証型広告」を三位一体で推進することを発表いたしました。
インターネットに占めるボットのトラフィック数が53%。
AIエージェントによるトラフィック数も前年比で7,851%となっています。
こうしたなか、広告業界では、ボットなどを使って広告の表示やクリックを偽装し、広告費を不正に詐取する「アドフラウド」の被害が拡大しており、全世界で13兆円(2023年)。日本国内だけでも1592億円(2025年)の被害が発生するなど、実在する人間に広告を届けることが重要になりつつあります。
このようななか、今回の業務提携では、Ads for HumanityがWorld IDを活用し、「実在する人間」のみに広告を届けるソリューションを提供するとともに、人間認証型広告をAds for Humanity、LGエレクトロニクスと共に推進してまいります。
なお、「人間認証型広告」を先行導入したローソンエンタテイメント様においては、クリック率が3倍、1クリックあたりの広告費(CPC)が61%削減するなどの効果も本日発表されました。
こちらに関する詳細のプレスリリースも発表致しましたので、ご一読ください。
prtimes.jp/main/html/rd/p/00…
@worldnetwork
Ajay retweeted
Hot take: you should be able to delegate your @world_id to your personal agent.
@alexblania explains on @tbpn
In Tokyo 🇯🇵 right now and learning about the very frequent cyber attacks that are happening here.
A stolen password or convincing help-desk call shouldn’t be enough to take over an account.
@world_id’s approach to enterprise MFA: bind the account to the human, then verify that same human at login and recovery.
Authenticate the human not just the credential.
#サイバー攻撃 #CyberSecurity






















