Around 2024, in Qingdao, Shandong, a victim unfamiliar with cryptocurrency operations placed trust in an acquaintance—entrusting Zhang to open a Bitcoin account on a trading platform and wallet, and to assist in managing the assets. During the registration process, Zhang naturally acquired access to the wallet mnemonic phrase and account access permissions, subsequently transferring all 107 Bitcoins from the victim's account to a wallet address under his control, and gradually converting them into RMB, realizing about 660,000 yuan. At the time of the incident or based on the current market price, these 107 Bitcoins amounted to over 53.55 million yuan. The first-instance court in Qingdao ultimately deemed his actions constituted theft, with particularly huge amounts misappropriated, sentencing him to 10 years and 9 months of imprisonment and a fine of 100,000 yuan. On a similar timeline, in 2025, the U.S. federal prosecution turned its attention to another kind of "exploitation": Google software engineer Michele Spagnuolo was accused of abusing confidential, unpublished data like internal Google search trends, betting on the prediction market protocol Polymarket deployed on the blockchain through an account named "AlphaRaccoon," allegedly profiting over 1.2 million dollars, and was charged with commodity fraud, wire fraud, and money laundering; the case is still ongoing in judicial proceedings without a final verdict. On one side, the failure to guard the mnemonic phrase allowed Bitcoins, supposed to be under the control of private keys, to change hands easily on-chain; on the other side, confidential data was repurposed for on-chain markets, transferring information asymmetry from the real world into publicly transparent contracts. The two cases occurring in China and the U.S., led respectively by local courts and federal prosecution, point to the same fundamental issue: the so-called "trustless" crypto world simply offloads the trust of protocol layers to code, without eliminating the risks of interpersonal trust and information asymmetry.
Acquaintance Custody Turns Into Gun Theft: Qingdao's 107 Bitcoins Lost
In the Qingdao case, the story began with what seemed to be an ordinary "help registering an account." The victim knew nothing about Bitcoin operations, so they entrusted two critical tasks to the acquaintance Zhang: first, to open a Bitcoin account on the trading platform and wallet, and second, for Zhang to safeguard the associated mnemonic phrase and access permissions. In other words, from the beginning, this wallet's "life and death power"—the mnemonic corresponding to the private key—was not in the asset owner's hands but rather in the custodian's pocket.
Under the custodial status, Zhang used the mnemonic phrase he possessed to transfer the entire 107 Bitcoins from the victim's account to a wallet address under his control. On-chain, this was merely a momentary transfer of ownership; offline, he gradually converted part of it into fiat currency through trading platforms or over-the-counter channels, totaling about 660,000 yuan. Based on this and the market price at the time of the incident, the first-instance court determined that the 107 Bitcoins transferred equated to over 53.55 million yuan, categorizing his actions as theft of a particularly huge amount, sentencing him to 10 years and 9 months imprisonment, with a 100,000 yuan fine. Facing the charges, Zhang argued that he was performing a "protective takeover," but the court noted his repeated secret transfers and phased cash-outs were clearly inconsistent with mere asset protection, reflecting his subjective intent of illegal possession. More glaring was the disparity in numbers: while the stolen Bitcoins were worth over 50 million yuan at market price, only over 600,000 yuan actually landed in his pocket, exposing the gap caused by Bitcoin's violent price fluctuations and the challenges of realizing value. This also indicated that once the mnemonic phrase is handed over, the parties would not only struggle to notice asset loss in time, but also find it difficult to remedy the transfer that had already occurred on-chain through technical means.
Mnemonic Phrase is the Vault Key: Where is the Trust Boundary
Technically, Bitcoin and similar public chain assets do not "recognize people," only private keys. Common wallets use a string of seemingly random mnemonic phrases to derive a full set of private keys and address groups; whoever possesses this mnemonic phrase can initiate transfers on-chain, moving away all assets from the corresponding address, and once it is packed onto the chain, reversing the transfer is virtually impossible. Giving the mnemonic phrase to others, taking a photo to send it, or even casually screenshotting and storing it in chat records or cloud albums effectively equals duplicating a vault key that can be used at any time, though you may not know when that key will be used.
The uniqueness of the Qingdao case lies in that the victim, unfamiliar with operations, entrusted the opening of accounts and wallet management entirely to acquaintance Zhang from the start, and had long been without access to their own mnemonic phrase; thus, the actual control over the 107 Bitcoins was entirely in the other party's hands from a technical perspective. When handling the case, judicial authorities acknowledged that Bitcoin and other crypto-assets possess certain property attributes and can serve as objects in criminal cases. Conversely, in the relationship's characterization, the parties were defined as being based on trust and delegation rather than joint investment or lending, which means that Zhang's role in controlling the mnemonic and account permissions is closer to that of a "custodian" in the traditional sense of holding others' property. The tension revealed by the case is that although on-chain systems are decentralized, in the real world, you either properly safeguard this string of mnemonic phrases and bear the risks of operational errors and forgetfulness alone, or you hand over the key to others or institutions, thereby taking on the risk related to whether the custodian "will use the key." What truly needs visibility and definition is not the technology itself, but this invisible boundary of trust.
Google Engineer Uses Search Secrets to Access Polymarket
On the other side of the ocean, the "key" in hand is not the mnemonic phrase but search data. Michele Spagnuolo is a software engineer at Google, whose daily work allows him to see what ordinary people may never access: aggregated internal search trends, hot keyword temperature curves—these are regarded as confidential, unpublished information that can outline public sentiment and attention trends before real-world events occur.
In 2025, the U.S. federal prosecution charged him, claiming he did not lock this data behind the corporate firewall but brought it to the blockchain prediction market Polymarket: here, all betting records are publicly recorded on-chain, but the identity of the bettors and the underlying information they rely on are not automatically disclosed. It is alleged that he used an account named "AlphaRaccoon," specifically targeting contracts related to highly trending real-world events, accumulating profits exceeding 1.2 million dollars. The prosecution's logic for the charges of commodity fraud, wire fraud and money laundering is that he converted confidential information that should have been sealed within the company into a competitive advantage in the public market, which closely resembles traditional finance's path of profiting by abusing undisclosed significant information. However, whether this behavior of exploiting company data for betting on the blockchain prediction market will eventually be clearly categorized by courts and regulators as "insider information" remains to be determined in subsequent legal proceedings.
On-Chain Fairness Meets Off-Chain Insider: How Pure Can Prediction Markets Be
Decentralized prediction markets like Polymarket have a very simple selling point: protocols deployed on the blockchain, rigid contract rules, allowing anyone to bet on real-world events, with all transaction, holding, and clearing records publicly accessible on-chain. Visually, it resembles a completely even playing table—no market makers manipulating behind the scenes, no matchmaking engine "favoring" high-frequency players; all participants face the same odds and settling logic, with "on-chain disclosure and fairness" almost advertised on the product's banner.
However, the Spagnuolo case brings a deliberately overlooked issue to the forefront: when some participants possess confidential, unpublished data such as search trends internal to tech giants like Google, this seemingly flat playing surface is effectively bent by information advantages in the offline world. The on-chain contracts themselves are not the problem, and all betting paths are transparent; what remains opaque is the information sources behind the traders—you can only see how much "AlphaRaccoon" bet in which event market, but you cannot see why he dared to heavily invest in a certain outcome while others hesitated. Unlike traditional securities markets, where the boundaries around "insider trading" have been relatively defined over the years through case law, the regulatory attribute of prediction markets remains contentious. The U.S. often applies existing charges like commodity fraud and wire fraud to file cases; whether the exploitation of non-public data for betting is directly categorized under "insider trading" still depends on future judicial rulings. At this uncertain boundary, protocol project parties must consider how to respond to regulatory inquiries regarding unusual accounts and suspicious patterns, companies holding vast amounts of data need to more explicitly restrict "taking confidential data to bet on-chain" in their employee codes, and ordinary participants must also become aware that so-called "on-chain fairness" can only cover the contracts and transaction layer itself; the legal and compliance constraints regarding the information brought into this market are becoming new gatekeepers.
From Bitcoin Wallets to Prediction Markets: How Individuals Can Protect Themselves
From the Qingdao case to the Google engineer case, while the technical architecture can be highly decentralized, the real determinants of risk direction remain human: on one end, handing over the mnemonic phrase to "someone more knowledgeable" results in 107 Bitcoins, according to AiCoin data, changing hands instantly on-chain, with legal remedies lagging and expensive; on the other end, internal personnel holding vast amounts of confidential search data bring the information advantage into the on-chain prediction market, magnifying profits, causing the so-called "public and transparent" contracts to be entirely rewritten by asymmetric cognitive frameworks. For ordinary participants, the foundational self-protection principles are not complicated: mnemonic phrases and private keys should never be entrusted to any third party, even if they are family or long-time acquaintances; any arrangement that requires you to "hand over the account" or "let them operate on your behalf" intrinsically relinquishes control; when faced with opportunities marketed as "stable high returns" or "risk-free profits," the first reaction should not be excitement but a critical inquiry into the source of funding and information advantages. Beyond the individual, large tech companies possessing vast data must establish clearer lines in internal compliance and access control, explicitly prohibiting employees from migrating confidential data to on-chain betting; protocols deployed on the blockchain also can no longer completely distance themselves by claiming "just a tool," but must provide more responsible answers within the boundaries of risk warnings, user education, and cooperating with judicial investigations. The two cases disclosed between 2024 and 2025, entering criminal procedures in China and the U.S. respectively, indicate that future similar cases will directly impact individuals' understanding of the true meaning of self-custody, their perspective on the boundaries of leveraging information advantages in decentralized applications, and their reevaluation of the risk costs associated with each "on-chain" action under increasingly dense judicial and regulatory scrutiny.
Join our community, let's discuss and become stronger together!
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata
Exclusive Hyperliquid benefits from AiCoin: https://app.hyperliquid.xyz/join/AICOIN88
Exclusive Aster benefits from AiCoin: https://www.asterdex.com/zh-CN/referral/9C50e2
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
