Crypto.com Spends $70 Million to Acquire AI.com, ERC-8004 Boosts AI Agent Infrastructure

marsbit2026-02-09 tarihinde yayınlandı2026-02-09 tarihinde güncellendi

Özet

Crypto.com CEO Kris Marszalek has acquired the domain AI.com for $70 million, marking the largest domain purchase in history and signaling a strategic shift from serving human users to building infrastructure for AI agents. This move aligns with the emerging vision of Decentralized AI Finance (DeFAI), where autonomous AI agents—not humans—become primary financial users. Central to this transition is the ERC-8004 protocol, which enables AI agents to have verifiable on-chain identities and reputation systems. This allows AI to operate independently with its own wallet and credentials, rather than acting merely as a human-controlled tool. Crypto.com aims to position AI.com as a gateway for AI agents to access efficient, programmable financial services. Key challenges addressed include payments—where stablecoins offer 24/7 instant settlement suited to AI’s speed—and identity, where decentralized identifiers (DIDs) and zero-knowledge proofs enable trust without central intermediaries. This shift reflects a broader industry transition toward “Agentic Finance,” where AI-driven transactions could dominate market liquidity. If successful, Crypto.com may redefine crypto exchange competition by capturing the emerging AI-agent economy, potentially marginalizing traditional human-focused platforms. The $70 million investment is a bet on this future, where AI agents drive financial activity through semantic understanding and algorithmic execution.

Author: Max.S

Crypto.com CEO Kris Marszalek plans to play a $70 million trump card during the Super Bowl's halftime advertisement — AI.com.

This news has swept through the financial and tech circles in the past 48 hours. This is not only a record-breaking domain name acquisition (surpassing the $30 million record of Voice.com) but also a highly symbolic "declaration": this crypto giant, which once built its fame by sponsoring stadiums and inviting Hollywood stars, is now redirecting its massive capital and traffic machine from serving "human retail investors" to fully serving "AI agents."

If past crypto bull markets were driven by human greed and fear, then Kris Marszalek is betting that the next cycle will be dominated by algorithms, code, and autonomous agents.

This is not just a rebranding; it is the "Normandy landing" of the DeFAI (Decentralized AI Finance) era.

Over the past fifteen years, all infrastructure in the cryptocurrency industry — from the mnemonic phrase user experience of wallets to the candlestick chart interfaces of exchanges — has been designed for humans. We discussed how to make it easier for "people" to use USDC for payments and how to help "people" understand complex DeFi yields.

However, the acquisition of AI.com opens a new chapter for the industry: the primary financial users of the future may not be human at all.

According to the latest disclosures from Forbes and The Block, Crypto.com plans to use AI.com to build a "decentralized self-evolving agent network." This is a critical signal. Traditional Web2 AI models (such as OpenAI or Google's Gemini) are centralized and siloed; meanwhile, the Web3 narrative is shifting toward granting these AI agents "sovereignty."

Here, we must mention a new standard that has sparked heated discussions in tech circles — ERC-8004. Just last month, the Ethereum community began intensive discussions about this protocol, aimed at endowing AI agents with verifiable identity and reputation systems on-chain.

Why is this important?

In the Web2 world, if an AI Agent wants to book a hotel or buy stocks, it needs to bind to an individual's bank card to pass human identity verification. At this point, the AI Agent is merely a "subsidiary" of humans. But in the Web3 vision, through protocols like ERC-8004, an AI Agent can have its own "Soulbound Token" (SBT) as an ID and its own wallet address. It is no longer a co-pilot; it is the pilot.

Crypto.com's purchase of AI.com is not just to sell a chatbot similar to ChatGPT. Its ambition is to become the entry point for these hundreds of millions of "silicon-based financial users." When your personal AI assistant needs to allocate funds, conduct arbitrage, or pay service fees via API, it requires an extremely efficient, low-friction, and compliant financial layer. By controlling the premium entry point of AI.com, Crypto.com aims to define the standards for this layer.

In the DeFAI architecture, two core issues — payment and identity — have not been perfectly solved by Web2 giants, and this is precisely the killer application of crypto technology.

Payment Problem:

Why must AI use cryptocurrency? Imagine a scenario: an AI Agent responsible for optimizing your investment portfolio spots a fleeting arbitrage opportunity. If it uses the traditional banking system, it must deal with T+1 settlement cycles, cumbersome cross-border remittance verifications, and potential risk control freezes.

AI operates at millisecond speeds; it cannot tolerate the inefficiency of "carbon-based finance." Stablecoins are the native currency of AI. They are online 24/7, programmable, and settle instantly. Crypto.com's full payment of $70 million in cryptocurrency to acquire the domain name is itself a massive "live demonstration." In the future, collaboration between AI Agents — for example, a "data analysis Agent" purchasing data from a "storage Agent" — will be completed entirely through on-chain micropayments.

Identity Problem:

How do you trust a counterparty without a physical body? This is the most fascinating and dangerous aspect of DeFAI. On the internet, you don't know if the other party is a dog; in the future metaverse, you won't know if the other party is human or AI, or even malicious code disguised as a benevolent AI. Traditional OAuth logins (e.g., "Sign in with Google") hand control to big corporations. In contrast, crypto technology provides DID (Decentralized Identity) and Zero-Knowledge Proofs (ZK-Proofs), allowing AI Agents to prove the following without relying on centralized servers:

"I was generated by a certain audited codebase."

"I possess sufficient computing power or funds to pay for this service."

"My past behavior has been good (based on on-chain reputation)."

The emergence of protocols like ERC-8004 is precisely to build this permissionless trust. Crypto.com's strategy suggests they aim not only to be an exchange but also to become the "identity registry" and "clearinghouse" for AI Agents.

Spending $70 million on a domain name may seem insane from a traditional finance perspective, but in the context of the crypto industry, it is the ultimate harvest of the "attention economy."

With the proliferation of LLMs (Large Language Models), the entry points of the internet are undergoing a fundamental shift. Users no longer go to Google to search; they ask AI directly. Users no longer manually place orders on exchanges; they let AI configure their assets. Whoever controls the AI dialog box controls the distribution of traffic.

Traditional crypto exchanges (especially CEXs) are facing severe homogenized competition. Fee wars have hit the bottom, and the effect of listing new coins is diminishing. Kris Marszalek keenly realizes that if most future trading instructions come from AI, then traditional app interfaces will become less important. What matters is API accessibility and the brand trust of the domain name.

AI.com is a self-explanatory brand. For "outsiders" trying to enter the Web3 world, it is the most intuitive entry point. For AI Agents, it could become a super aggregator that invokes complex financial services through natural language instructions.

This is a competition about survival. If Coinbase and Binance remain stuck in the mindset of serving "human traders," while Crypto.com successfully transforms into a "financial service provider for AI," then in the next bull market dominated by machine liquidity, existing giants could be dimensionally reduced, much like Nokia in its day.

If Crypto.com successfully redefines the market, DeFAI will experience explosive flywheel growth: more AI Agents using cryptocurrency -> increased cryptocurrency liquidity -> more robust infrastructure -> attracting more traditional AI companies to join. This is a positive feedback loop. The $70 million gamble is betting on the start of this flywheel.

Human traders are driven by greed and fear, which creates market volatility and the possibility of profiting by "harvesting emotions." But AI Agents are cold and calculating. They trade based on data, probabilities, and preset utility functions, which could lead to markets becoming extremely active and efficient. Alpha returns will become exceedingly hard to find unless you have an absolute advantage in algorithms or information acquisition speed. The survival space for retail investors will be further squeezed, forcing them to entrust their funds to AI Agents for management.

Future trading will not only be about the exchange of numbers but also the exchange of semantics. An AI Agent might read a news article about Federal Reserve policy, understand its semantics, and adjust its holdings in milliseconds. AI.com could become the command tower for this "semantic finance," allowing ordinary users to simply input a command like "help me hedge risks based on the latest macro situation," and the underlying Agent would automatically call on-chain protocols to execute the operation.

Crypto.com's acquisition of AI.com, in the early spring of 2026, may look like an expensive marketing stunt. But if we extend the timeline to ten years, this might be seen as a watershed moment in cryptocurrency history. It marks our transition from "Internet Finance" (FinTech) to "Agentic Finance."

The outcome of this gamble is still unknown, but Crypto.com has already pushed its chips to the center of the table. As Kris Marszalek implied: in the face of the AI tidal wave, you either become its infrastructure or become its obsolete footnote.

İlgili Sorular

QWhat is the significance of Crypto.com's $70 million acquisition of AI.com, according to the article?

AThe acquisition is a strategic move to pivot from serving human retail users to becoming a financial services provider for AI agents, positioning Crypto.com as a key entry point for 'silicon-based financial users' and aiming to define the standard for decentralized AI finance (DeFAI).

QWhat is ERC-8004, and why is it important in the context of AI agents and Web3?

AERC-8004 is an Ethereum protocol proposal that provides AI agents with a verifiable on-chain identity and reputation system, allowing them to have their own wallet addresses and act autonomously rather than as human appendages, which is crucial for trust and decentralization in Web3.

QHow does the article contrast Web2 and Web3 approaches to AI agent identity and payment systems?

AWeb2 AI agents rely on centralized systems like OAuth and human-linked bank cards, making them dependent on big companies and slow traditional finance. Web3 uses decentralized identity (DID), zero-knowledge proofs, and stablecoins for instant, programmable, and permissionless transactions, enabling AI agents to operate independently and efficiently.

QWhat are the two core problems in DeFAI (Decentralized AI Finance) that the article highlights, and how does crypto technology address them?

AThe two core problems are payment and identity. Crypto technology addresses payment with stablecoins for 24/7, instant settlements and addresses identity with protocols like ERC-8004 for verifiable, decentralized identity and reputation without relying on central authorities.

QWhat long-term vision does the article suggest Crypto.com has with the AI.com acquisition, and what potential impact could it have on the crypto industry?

ACrypto.com aims to become a central hub for AI agent-driven financial services, potentially triggering a flywheel effect where more agents use crypto, increasing liquidity and infrastructure development. This could shift the market from human-driven trading to AI-dominated 'Agentic Finance,' squeezing out散户 and demanding algorithmic superiority for alpha generation.

İlgili Okumalar

Can Generative Models Finally Be Trained End-to-End? The Core Is a For Loop

This article introduces a novel training paradigm for generative models called Explorative Modeling (XM), which enables true end-to-end training. Traditionally, powerful generative models like autoregressive and diffusion models are not trained end-to-end. They are trained to predict a single small step but require iterative multi-step sampling for inference. This "exposure bias" leads to error accumulation and limits performance. The core challenge XM addresses is "mode blurring." In generative tasks, a single input (e.g., "generate a dog") corresponds to many valid outputs (multiple modes). Standard training objectives like reconstruction loss force the model to average these modes, producing unrealistic, blurry outputs. To avoid this, existing models break generation into many small, almost deterministic steps, sacrificing end-to-end training. XM tackles this by restructuring the training loop itself. Its key insight is to amplify "generative expressivity." For each training input, instead of generating one sample, the model generates K candidate outputs. Only the candidate closest to the real data is used for computing the loss and updating the model via backpropagation. This simple "best-of-K" mechanism is implemented as a short for-loop. By exploring multiple possibilities, the model learns to distribute its guesses across different modes rather than collapsing to their uninformative average. The paper demonstrates that "exploration" acts as a new, powerful scaling axis. Gains from XM increase with model size, data scale, and compute. Experiments show improvements in FID scores for image generation and significant efficiency gains, sometimes outperforming larger models without exploration. When pushed to the limit, XM enables fully single-step, end-to-end generative models. In robotics tasks, an "Explorative Policy" matched the performance of a 100-step Diffusion Policy with a single forward pass, drastically improving inference speed. While the best-of-K concept is not entirely new, the authors' contribution lies in formally understanding it as a direct method to boost generative expressivity without fragmenting the generation process. This work suggests that as models scale, enhancing exploration during training may become crucial for overcoming fundamental performance bottlenecks.

marsbit49 dk önce

Can Generative Models Finally Be Trained End-to-End? The Core Is a For Loop

marsbit49 dk önce

Annual Salary of Millions Competing for Electricians, Meta Rushes to Open Its Own Technical School

The AI boom is facing an unexpected bottleneck: a severe shortage of skilled construction workers and electricians. As tech giants like Meta, OpenAI, and Alphabet race to build massive data centers—such as OpenAI's $16 billion "Stargate" project—they are hitting a critical labor wall. The U.S. needs an estimated 130,000 more electricians, 240,000 construction workers, and 150,000 supervisors by 2030 for AI infrastructure alone, but tens of thousands of electrician jobs go unfilled each year. While AI companies offer high premiums, with electricians earning up to $280,000 annually, worker scarcity still causes massive losses—delays on a single project can cost $14.2 million per month. The complexity of building AI data centers, which require immense power (equivalent to powering hundreds of thousands of homes), sophisticated electrical systems, and advanced liquid cooling solutions, demands highly skilled technicians who are in short supply. To combat this, companies are investing heavily in training. Meta has committed $115 million to a free training school offering tuition, housing, and stipends, targeting 5,000 new workers. OpenAI is partnering with unions to secure skilled labor. These efforts are paying off, with a significant rise in Gen Z interest in trade schools over college. However, the power demands are staggering. AI data centers are driving a rapid surge in electricity consumption, projected to account for up to 12% of U.S. power use by 2028 and raising costs for consumers. Furthermore, the construction boom is project-based, leading to a potential future glut of trained workers once building peaks, which could depress wages industry-wide. The race for AI supremacy now depends as much on skilled hands as on advanced chips.

marsbit2 saat önce

Annual Salary of Millions Competing for Electricians, Meta Rushes to Open Its Own Technical School

marsbit2 saat önce

İşlemler

Spot
活动图片