Binance Founder CZ Says AI Can Cause Damage Without Breaking Bitcoin's Cryptography
Binance founder Changpeng Zhao stated on social media that while people are concerned about AI attacks on crypto assets, the encryption strength relied upon by the internet and banking systems is weaker.
He mentioned that AI does not need to actually break Bitcoin's cryptography to cause real damage. Leaked data is sufficient to support larger-scale phishing and social engineering attacks.
He concluded with "Stay SAFU." This statement comes in the context of the disclosure of a critical PHP vulnerability, after which CZ had previously warned users to be cautious when accessing potentially unpatched websites.
Cryptographic systems generally use stronger algorithms like elliptic curves, while many traditional websites and banking systems still use older encryption standards. Attackers can increase the success rate of phishing by combining data leaks with AI-generated realistic information.
Market Mechanism: Attackers use leaked data and AI-generated content to lower the cost of phishing; users bear the risk of losing private keys and accounts; exchanges and wallet providers need to enhance user education and detection tools; traditional internet and banking systems with weaker encryption become priority targets; the main exposure point for crypto asset users shifts from cryptography itself to human operations and data leaks.
Source: Public Information
ABAB AI Insight
Changpeng Zhao has long held roles related to Binance and has publicly spoken multiple times about user security and system risks. This statement distinguishes concerns regarding AI and cryptography.
In terms of capital pathways, data leaks themselves constitute tradable or usable assets, and AI tools reduce the marginal cost of transforming leaked information into personalized phishing content. User funds are transferred through social engineering rather than password cracking.
This is similar to previous types of crypto phishing, address poisoning, and deep forgery incidents, and is close to the model of fraud in traditional finance using leaked data. We are currently in a phase where data leaks are normalized and AI tools are widespread.
Essentially, this represents a transfer of pricing power: the cost of attacks shifts from breaking strong cryptography to utilizing existing data and generating content. The mechanism is that strong encryption raises the direct cracking threshold, but human decision-making and exposed information become lower-cost entry points; AI amplifies the efficiency of information utilization, increasing the success rate and scale of phishing and social engineering without needing to break the underlying mathematics.
ABAB News · Cognitive Law
AI does not need to break cryptography; leaked data is sufficient to initiate phishing.
The encryption of the internet and banks is weaker; the cryptography of crypto assets is not the weakest link.
The cost of social engineering decreases due to AI, making user operations the main exposure point.