Cross-chain security issues remain a major pain point for crypto markets, after investigators reported two separate bridge-related exploits occurring only hours apart. According to on-chain analytics firm Blockaid, the combined theft totaled more than $31.6 million, with funds taken from bridge infrastructure used by decentralized perpetual exchange AFX and the Verus Ethereum Bridge.

Blockaid said AFX’s bridge lost $24.15 million on Wednesday, before another attack targeting the Verus Ethereum Bridge resulted in roughly $7.5 million drained from bridge reserves. The back-to-back incidents underscore how bridge operators—and the protocols that integrate them—can be exposed even when exploits are not tied to a single chain-level weakness.

Key takeaways

• Blockaid reported losses of $24.15 million from an AFX-operated bridge on Arbitrum and about $7.5 million drained from the Verus Ethereum Bridge within hours.
• Offchain Labs co-founder Stephen Goldfeder said Arbitrum’s native bridge was not hacked, pointing to activity originating from a third-party protocol.
• Security researchers suggested the AFX incident may have involved compromised keys rather than a smart contract logic flaw.
• Blockaid said the Verus exploit appears to mirror a prior May incident, using a similar method while involving a different attacker wallet.
• Both cases highlight that bridges remain high-value targets because they custody large asset pools and move value across ecosystems.

AFX bridge exploit on Arbitrum: what was targeted

Blockaid said it detected an exploit at 9:30 pm UTC aimed at a bridge operated by AFX, a decentralized perpetual exchange running on Arbitrum. The investigation framed the event as a bridge compromise affecting a third-party integration rather than a breach of Arbitrum’s core bridging infrastructure.

According to Offchain Labs co-founder Stephen Goldfeder, a bridge hack report circulating online had impacted a transaction originating from a third-party protocol, and that the Arbitrum native bridge itself had not been exploited. Goldfeder stated that the transaction in question originated from another protocol and emphasized that Arbitrum’s native bridge “has not been hacked or exploited in any way.”

Additional analysis from SunSec, the founder of the DeFi security community DeFiHackLabs and a contributor to SEAL, suggested that the evidence pointed more toward compromised keys than toward a vulnerability in smart contract logic. While that distinction matters for incident response—key compromise typically demands urgent credential rotation and broader access review—it also signals that the weakest point may not always be the bridge contracts themselves.

Cointelegraph sought comment from AFX regarding the reported exploit, but the additional reporting available here centers on what Blockaid and affiliated investigators observed during the incident.

Verus Ethereum Bridge attack: a similar method to May

In a separate incident, Blockaid reported an exploit targeting the Verus Ethereum Bridge that drained approximately $7.5 million across multiple assets held in bridge reserves. The listed tokens included Ether (ETH), tBTC, USDC, USDt, EURC, MKR, and scrvUSD.

Blockaid said the attack method appears similar to a previous Verus Ethereum Bridge incident reported in May, which resulted in the theft of $11.58 million. In that earlier case, Blockaid said the same overall approach was used, but by a different attacker wallet.

According to Blockaid, the attacker used the bridge “import path” to trigger “unbacked Ethereum-side payouts.” In practical terms, this points to a workflow-level weakness: attackers may be able to induce the bridge to release assets on one side of the system without corresponding backing on the other side, creating a direct path to reserve depletion.

For users and integrators, the repeated nature of the tactic raises a persistent risk: even when teams patch one vulnerability, the operational mechanics of how imports and payouts are handled can remain exploitable if the underlying assumptions aren’t fully addressed.

Why bridge failures keep recurring

Bridge exploits are difficult to eliminate entirely because cross-chain infrastructure often combines multiple components: custody of assets, message passing or import/export mechanisms, and permissioning for triggering settlement flows. When attackers find a seam between those elements—whether through compromised credentials, incorrect authorization, or weaknesses in how cross-chain states are validated—the result is frequently rapid draining of funds.

On-chain investigator TheCrypticWolf summarized the broader issue in a post on X, arguing that bridges remain a weak link until “security is upgraded.” While that statement reflects a general view rather than new incident-specific evidence, the two reported attacks within the same day give it concrete support: high-value bridge reserves make the system attractive, and high complexity makes comprehensive hardening challenging.

There is also an important asymmetry across the two incidents. Blockaid’s reporting on the AFX case was paired with Goldfeder’s clarification that Arbitrum’s native bridge was not compromised, suggesting the problem lay in third-party integration or bridge controls tied to a particular protocol. In contrast, Blockaid’s description of the Verus incident emphasizes how the bridge import mechanism can lead to Ethereum-side payouts that are not properly backed—an issue that may relate more directly to settlement logic and state assumptions.

What to watch next for affected ecosystems

Bridge-related incidents typically lead to emergency measures such as pause controls, increased monitoring, and changes to custody or authorization workflows. Readers should watch for follow-up disclosures from AFX and the Verus ecosystem, especially around what Blockaid and other investigators determine about root cause—whether it’s key compromise, an authorization failure, or a repeatable weakness in import/export settlement.

More broadly, these events reinforce that cross-chain exposure isn’t limited to the bridge operators alone: decentralized applications and traders relying on bridges for liquidity and settlement should treat bridge security as a continuously evolving risk, not a one-time checkbox.

Circle has signed separate memorandums of understanding with Kakao Group and South Korean fintech operator Toss to explore stablecoin payments, blockchain settlement and digital asset infrastructure in South Korea.

The agreements bring Circle’s USDC and payment technology into discussions with some of Korea’s largest consumer finance platforms. Kakao, Kakao Pay and Kakao Bank will study opportunities around KRW-based digital assets, cross-border payments and tokenized financial services. Toss and Toss Bank will examine similar uses, including digital wallets, overseas payments and programmable onchain transactions.

Kakao Group said its agreement with Circle will combine the KakaoTalk-centered platform ecosystem with Kakao Pay’s payment services, Kakao Bank’s banking capabilities and Circle’s blockchain infrastructure. The companies plan to review payment, settlement and digital asset connectivity as South Korea develops rules for stablecoins and other tokenized financial products.

Another week, another multi-million-dollar hack in DeFi, and once again, it’s an off-chain compromise rather than a smart contract exploit.

AFX Trade, a decentralized perpetuals exchange that settles in dollar-pegged stablecoin USDC, was drained of about $24.15 million on Wednesday after an attacker compromised the validator signing keys behind a bridge the protocol operates on Arbitrum, blockchain data shows.

In other words, the smart contract did what it’s supposed to do – verify the signature and execute the transaction. The problem was with the private keys that generated those signatures, as attackers compromised the private validator signing keys (hot keys held offchain by the bridge operators or validators).

Steven Goldfeder, co-founder of Offchain Labs, which develops and maintains the network, said the Arbitrum native bridge “has not been hacked or exploited in any way” and that the transaction originated from a third-party protocol.

A hack of Arbitrum’s own bridge would signal risk across the entire layer-2 network, but a compromised protocol running on top of it is a contained failure.

Nothing in the bridge’s own code logic was broken. Bridges are blockchain-based tools for transferring tokens between various networks, including those they were not initially supported on.

A senior official from the White House’s Office of Science and Technology Policy has accused the Chinese AI firm behind Kimi K3 of using “covert industrial distillation” techniques to replicate capabilities from U.S. models. The allegation, posted to X on Wednesday by Michael Kratsios, underscores how U.S. concerns about AI competitiveness are increasingly blending with fears of large-scale intellectual property (IP) theft.

Kratsios said the company built an internal platform to distill U.S. models “at scale,” specifically using methods intended to evade detection. While he argued that distillation—compressing a model into a smaller one—can be legitimate and part of open innovation, he framed the alleged approach as unacceptable because it targets proprietary American technology rather than improving models through transparent research.

Key takeaways

• White House OSTP Director Michael Kratsios alleged Chinese firm Moonshot AI used large-scale covert distillation tied to the Kimi K3 release.
• Kratsios contrasted legitimate model distillation with alleged industrial-scale techniques aimed at stealing U.S. IP and avoiding detection.
• Some AI researchers dispute claims that Anthropic’s Fable was used to produce Kimi K3’s performance, citing technical plausibility and timing constraints.
• U.S. officials warned that sanctions and Entity List designations could follow IP-theft-style distillation attacks.

Why the allegation matters beyond headlines

AI distillation is not inherently controversial. In general terms, distillation helps create smaller, more efficient models by training them on outputs generated by a larger “teacher” model. The White House’s argument, as stated by Kratsios, is that scale and secrecy change the nature of the activity—turning a common engineering practice into something closer to a targeted extraction of proprietary capability.

That distinction is critical for investors, developers, and researchers because it signals a potential shift in how regulators and governments may view certain AI training pipelines. If authorities treat “covert industrial distillation” as IP theft, it could influence enforcement priorities, compliance expectations, and the willingness of model providers to share weights, outputs, or licensing terms—especially across geopolitical lines.

Timing and the dispute over Anthropic’s role

Kratsios’s claim places particular focus on the question of whether U.S. model technology was used in the preparation of Kimi K3. Cointelegraph previously reported that Anthropic’s Fable 5 was taken offline quickly due to U.S. export controls, then re-released on July 1. Kimi K3, meanwhile, launched on July 16—creating what critics describe as a narrow window for any distillation-derived transfer.

Elie Bakouch, a researcher at Prime Intellect, publicly questioned whether the technical story matches the observed outcomes. In an X post referenced in the original reporting, Bakouch argued that there are only “15 days between fable 5 ban removal and kimi K3 release,” and he added that the performance “could” not be explained in a straightforward way by distillation from Fable.

Dean Ball, head of strategic futures at OpenAI, also pushed back. On Friday, Ball said he did not believe K3’s performance could be “explained away by distillation or anything like that.” Both responses reflect a broader point: even if distillation happened, it may not be the sole—or even the primary—reason for a model’s capabilities, and establishing a clean causal link can be technically difficult.

In the absence of publicly available technical evidence, these disputes matter because they highlight uncertainty. Government accusations may have intelligence backing, but for the wider AI community, the plausibility and traceability of model-to-model influence is a separate question from whether the activity would violate policy or law.

Washington escalates from concerns to potential enforcement

The posture from U.S. officials appears aimed at deterrence. In addition to Kratsios’s claim that “covert industrial distillation” intended to steal U.S. technology is unacceptable, U.S. Treasury Secretary Scott Bessent warned that sanctions and restrictions could be pursued.

Bessent said the U.S. supports open-source AI and the innovation it enables, but he argued open source does not mean “open season” on American IP. He also warned that if firms conduct covert, industrial-scale distillation attacks that cross into IP theft, consequences could include sanctions and Entity List designations.

That statement suggests the U.S. may attempt to treat certain distillation behaviors under the same enforcement logic used for other technology-transfer and IP-protection efforts. For AI companies, the practical takeaway is that even widely used ML techniques could be reinterpreted depending on intent, transparency, and scale.

It also raises a policy tension: distillation can improve accessibility and efficiency, but enforcement actions could push industry toward more restrictive handling of model outputs and training procedures. Developers may respond by tightening documentation, auditing data provenance, or changing how they handle third-party model access.

What to watch next

Whether the dispute becomes a broader enforcement campaign will likely depend on what additional evidence, if any, is made public and how regulators define “industrial-scale” and “covert” distillation in measurable terms. For now, observers should watch for any formal government actions tied to Kimi K3 and for further clarification from researchers on what technical signals can reliably connect teacher models to student performance.