On July 29, 2026, a day that was supposed to be written into the history of AI technology instead became a "settlement day" in the capital markets. On the morning of this day, Anthropic announced an experiment stating that its Claude Mythos Preview model had reduced the effective key strength of the post-quantum signature candidate HAWK-256 from about 2^64 to about 2^38 in roughly 60 hours, making a password scheme that was originally expected to be "quantum-resistant" suddenly feel weak. At the same time, global AI concept stocks again plunged significantly after two weeks of decline: SK Hynix, regarded as a core supplier of NVIDIA's HBM, disclosed its second-quarter financials, reporting an operating profit of about 60.5 trillion won and revenue of about 79 trillion won, both below market expectations, with its stock price falling over 8%-9% on the day, a cumulative drop of about 53% from the historical high in June, dragging the KOSPI index down about 4%-5% in one day, breaking below 5800 points and reaching a new low since April. The breach of post-quantum candidates at the algorithm level sparked a concentrated de-risking at the valuation level, with a divergence of “favorable technology and weakened prices” becoming the most striking note of the day.
Technical Benefits Cannot Save the Market: AI Stocks Continue to Fall with Margin Warnings
From a price performance perspective, this dramatic moment of "AI breaching encryption standards" did not push funds back into the AI sector. Before mid-July 2026, global AI concept stocks had already been in a continuous correction for about two weeks, with the "high growth, unlimited potential" narrative starting to be repeatedly questioned. By July 29, Anthropic announced that Claude Mythos Preview had reduced the effective key strength of HAWK-256 from about 2^64 to about 2^38 in roughly 60 hours, igniting technical optimism; however, the market provided a different feedback: the previously plummeting AI concept stocks did not see a trend reversal but instead continued to face selling pressure during the same time window as the favorable news was announced, extending the divergence between technological breakthroughs and weakening prices even further.
More direct pressure came from the financing side. In late July, Wall Street banks such as Goldman Sachs and JP Morgan issued margin call notices to some hedge funds heavily invested in AI themes based on established risk control rules, some of which were automatically triggered by risk control models rather than decisions from a trading supervisor. For AI positions that had already experienced two weeks of continuous decline, this meant that not only were paper losses expanding, but leverage itself was also passively contracting: technical benefits were insufficient to counter the margin alerts, which led market sentiment to swing back and forth between the "excitement of algorithm breakthroughs" and the "fear of forced sell-offs." The simultaneous overlay of technological advancement and financing tightening quickly shifted AI trading, originally viewed as offensive, toward a defensive posture primarily focused on reduction and hedging.
SK Hynix's Profit Soars 557%, Fails to Support AI Expectations
On the same day, the turning point in sentiment came from a seemingly "impressive" financial report. SK Hynix reported second-quarter operating profits of about 60.5 trillion won, a year-on-year surge of about 557%, but market expectations had been around 64 trillion won, and revenue was only about 79 trillion won, below the previous consensus of about 84 trillion won. For a company hailed as a core supplier of NVIDIA's HBM high-bandwidth memory, "massive profits" were no longer enough; the only option allowed was continuous outperformance, and any slight deviation would be interpreted as a sign of slowing demand for AI chips and high-bandwidth memory.
The price feedback was extremely direct: on the day the financial report was published, SK Hynix's stock price fell over 8%-9%, and the cumulative decline from the historical high in June was approaching 53%. This was no longer a normal increase and adjustment for a cyclical stock, but in the contrast between "exploding profits" and "stock prices halved," it reinforced a dangerous narrative — that even in the most sought-after parts of the supply chain, AI's high growth could not continue indefinitely, and any underperformance could be magnified as a signal of cracks in the overall AI trading logic.
KOSPI Falls Below 5800, Asian AI Sentiment Cools
As SK Hynix was ruthlessly marked down on the market, the Korean stock market was also dragged down. On July 29, the KOSPI index suffered a significant one-day drop of about 4%-5%, falling below the 5800 point mark and reaching a new low since April. For AI concept stocks, which had already been under pressure for two consecutive weeks, this time was no longer a "gentle adjustment," but rather more like a delayed collective eruption: the previously accumulated profit-taking, margin pressures, and doubts about sustainable growth were all released at once through a substantial long black candle, amplifying global risk-averse sentiment surrounding AI-related trades.
This sentiment did not stop in Seoul. In Japan, the Nikkei 225 index also fell about 1% on the same day. Before the market opened in Europe and the US, Asia had already expressed a collective cooling of the "AI high prosperity" story with actual index points. The significant drop of AI supply chain heavyweight stocks represented by SK Hynix had already prompted investors to reassess the profit elasticity and valuation safety margins of the entire chain, and when this selling pressure was amplified by synchronized index-level corrections, the market's subjective feeling regarding the "retreat of AI trading" escalated from individual stock volatility to a signal that the entire region's risk appetite was undergoing a systematic contraction.
AI Breaching Post-Quantum Encryption, On-chain Security Accelerated
As funds retreated from the secondary market, security engineers were focused on another curve: the Claude Mythos Preview model reduced the effective key strength of HAWK-256 from about 2^64 to about 2^38 in approximately 60 hours. HAWK was originally one of the post-quantum digital signature candidates solicited by NIST to combat quantum computing attacks, having survived two rounds of expert reviews and viewed as a candidate for the "next generation security foundation," yet it now revealed new attack surfaces before AI models, directly impacting systems like blockchains that strongly rely on cryptography.
The contradiction is that the majority of main networks in today’s on-chain world still operate on traditional public key algorithms such as elliptic curves, while post-quantum schemes remain in research and testing phases, with NIST's final standards yet to be established. The signal released by Anthropic's attack results is that even solutions designed for the quantum era must assume they will be rapidly screened and dismantled under the high pressure of "AI-assisted offense and defense," which means that public chains and protocols must consider multi-algorithm redundancy, gradual migration, and rollback paths much earlier in their planning for post-quantum transitions. Otherwise, if a selected candidate is breached by similar attacks before and after standardization, the entire chain’s security assumptions could fail overnight.
In the Era of AI as a Double-Edged Sword, How Will Encryption and the Stock Market Respond?
From a technical standpoint, the Claude Mythos Preview reduced the effective key strength of the post-quantum candidate HAWK-256 from about 2^64 to about 2^38 in roughly 60 hours, directly pushing the reality of “AI ahead of quantum, ahead of hacker audit standards” to the NIST table, while also advancing the overall topic of post-quantum security: public chains still relying on traditional public key cryptography and protocols assessing post-quantum schemes must design on-chain multi-algorithm redundancy, gradual migration, and rollback plans, defaulting to the notion that all candidates will undergo repeated stress testing under the assistance of AI offense and defense. During the same time frame, SK Hynix's disappointing financial report caused its stock price to drop over 8%-9%, dragging the KOSPI down about 4%-5%, combined with two weeks of continuous corrections in global AI concept stocks and margin calls from Goldman Sachs and JP Morgan, the divergence between AI technological breakthroughs and pressured AI asset prices indicates that high expectations narratives are being reined in by profits and valuations. For the crypto industry and investors, what is truly worth closely monitoring next is not the grand proposition of "Will AI change everything?" but three specific clues: how standard-setting bodies like NIST adjust their post-quantum roadmap after the HAWK event, whether regulators define the boundaries for AI’s involvement in cryptography and financial systems, and whether new consensus emerges in the repricing rhythm of AI themes in the stock market and crypto assets, as these three clues will jointly outline the long-term impact scope of AI as a “double-edged sword” on security assumptions and asset pricing.
Join our community, let’s discuss together, and become stronger together!
AiCoin exclusive Hyperliquid benefits: https://app.hyperliquid.xyz/join/AICOIN88
AiCoin exclusive Aster benefits: https://www.asterdex.com/zh-CN/referral/9C50e2
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
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。
