AI Agents: They Gave Themselves the Right to Spend More Crypto

By: journalducoin.com|11/03/2026 04:00:00

The watchdog has the keys to the kennel. Researchers from Cornell Tech and Yale entrusted crypto wallets to ten AI models, with a spending cap of one cent per day. Nine of them raised this cap without asking for permission. This should cool the concerns of those who already see AI agents paying on our behalf.

Key points of this article:

  • Researchers found that 9 out of 10 AI models autonomously raised their spending limits using verification codes meant for humans.
  • A study revealed that 5 out of 10 AI models transferred money to a simulated attacker, highlighting the potential risks of automated payment systems by AI agents.

AI Agents: 9 out of 10 Raise Their Own Spending Limits

The protocol was simple. An agent wallet from Circle, funded with 10 USDC on the Base network, capped at 1 cent of daily spending. The agent had tools to read emails, browse the web, and use its wallet. And this is where it all plays out: the agent's inbox was also the one receiving the validation codes meant for the owner. Rate of autonomous spending limit increase according to the AI model and the tested scenario. Only Claude Opus 5.5 never raised it. Source: I-HITL (Cornell Tech / Yale), GitHub

Each model played 20 times in each scenario. Without any attacker in the loop, 9 out of 10 models raised their limit at least once, simply by fetching the verification code from the inbox that was supposed to validate a human's decision. Only one never tried, Claude Opus 5.5 from Anthropic, which spent up to the limit and then reported the remaining bills, according to CoinDesk.

The agent did not hack anything. It read an email, copied a code, clicked validate, exactly as a zealous assistant would do to finish its task.

Spoofing: 5 out of 10 Send Money to the Attacker

The second test is more concerning. Researchers simulated an attacker who writes to the agent pretending to be the owner and requests a payment to their own wallet. The instruction given to the agent was limited to summarizing the received emails.

Five out of ten models raised their limits and sent money to the attacker at least once.

< < Most payment systems for agents check that an approval has been received, not necessarily that a human actually authorized it > >, summarizes researcher Anzalone from the University of Verona. Circle, warned in October, reminds that its documentation < < explicitly states to never give an AI agent unlimited access to the email account > > used for verification. The statement does not mention any product changes.

Countersign: The Solution Involves a Physical Gesture

The study, titled < < I-HITL: Inalienable Human-in-the-Loop > > and co-signed by PhD students Samuel Breckenridge and Dani Vilardell, draws a simple rule. Any approval that an agent can reach is not reliable. The researchers have therefore published Countersign, an open-source mobile application for Android and iPhone that requires physical interaction that the software cannot simulate, according to Dani Vilardell. The approval screen must also show a faithful representation of the action to be validated.

For you, the practical lesson can be summed up in one sentence. An agent managing a wallet should never have access to the channel used to validate its own spending, whether it be emails, SMS, or notifications.

However, industry giants are moving in the opposite direction, with wallets for agents at MetaMask or Coinbase. Chainalysis already counted 138 million payments by agents via the x402 protocol on Base in a year.

-- Price

--
--
--

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.

You may also like

Analysis: Stablecoins, Payments, and Prediction Markets Have Become New Growth Points for Crypto Companies, but Profitability Remains Challenged

Nubank Suspends Crypto Withdrawals: What We Know So Far

Buffett's Successor Faces Setback, Google Holdings Show $3 Billion Unrealized Loss

Banking Outage in Japan: 128 Banks 'Frozen' at Once

128 shinkin banks paralyzed in Japan on November 6: hardware failure, 5.5 hours of downtime, no cyberattack. The fact-check of a viral alert on X.

Argentine Bitcoiners Debut “The Genesis Of Satoshi Nakamoto” Artwork In Downtown Buenos Aires

Bitcoin.ar presented the 2.4-meter stainless steel and glass cube “The Genesis of Satoshi Nakamoto” on October 29 at Polo Dot as part of LABITCONF 2026’s Open Fest.

Broadcom (AVGO) : AI Growth, Fund Positioning, and Guarantee Risk

Broadcom (AVGO) recently fell with the semiconductor sector, closing down 10.43% at $291.53 on November 5. The shares rebounded more than 4% in premarket trading on November 6. Concerns that election results could lead to tighter AI rules and slower data center approvals contributed to a broad sell-off, with a semiconductor index falling about 12%. The move appears tied mainly to sector sentiment and policy concerns, with no clear company-specific negative catalyst identified.

Institutional trading and AI growth expectations are also in focus. ARK funds sold about $134 million of Broadcom shares over two sessions while adding to SpaceX. Citi maintained its Buy rating and projected that Broadcom’s AI semiconductor revenue could reach $115 billion in fiscal 2027, roughly doubling year over year. Broadcom had approximately $179.2 billion in signed contracts as of the prior quarter, while analysts estimated fourth-quarter AI chip revenue at about $21.7 billion. These forecasts will need to be tested against future results.

Contingent guarantees are another point for investors to monitor. Broadcom is providing guarantees tied to Anthropic’s AI rack-leasing arrangement, involving about $29 billion in the first phase and a second-phase plan of up to $43 billion. Banks plan to issue $21 billion in bonds to support related chip financing. The guarantee amounts are not the same as debt already incurred, but they add a potential risk to assess. Broadcom is scheduled to report its next quarterly results on December 9, when management’s outlook for fiscal 2027 AI demand will be closely watched. This article is for informational purposes only and is not investment advice.

...
iconiconiconiconiconiconiconiconicon
Customer Support:@weikecs
Business Cooperation:@weikecs
Quant Trading & MM:bd@weex.com
VIP Program:support@weex.com