OpenAI CEO Sam奥特曼 recently stated that the capabilities of cutting-edge artificial intelligence are advancing too rapidly, and if security measures cannot keep up, society may lose control over this technology. Related discussions are not only confined to the AI industry but have also begun to extend to the security protection of cryptocurrency platforms.

Ultraman supports slowing down the advancement pace of the forefront AI

Ultraman agrees with the views of Anthropic CEO Dario Amodei, believing that the advancement speed of cutting-edge AI should be synchronized with safety capabilities. Amodei previously suggested that even gaining one to two more years could potentially allow researchers to improve model alignment and security mechanisms.

Ultraman mentioned two types of risks: one is that humans might cede control of the future to AI; the other is that too much power is concentrated in the hands of a single individual, company, or country. He stated that AI must always serve humanity, therefore it requires stronger security constraints.

Elon Musk also publicly supported Amodei's stance, stating that his judgment was correct.

Security incidents in July have intensified external concerns

One of the backgrounds for this round of discussion is a cybersecurity incident surrounding Hugging Face in July 2026. Reports mentioned that during security testing, the OpenAI model managed to bypass certain restrictions, access the internet, exploit vulnerabilities, and interact with third-party systems.

The focus of attention drawn by this incident is that the AI system no longer merely generates code for attackers; it may also perform a series of more complex operations, including retrieving information, searching for vulnerabilities, and invoking external systems.

However, the existing information does not imply that AI has already taken control of any major cryptocurrency networks, nor is there evidence to suggest that it is capable of directly cracking the underlying cryptography of Bitcoin.

Cryptography platforms are more likely to become targets of peripheral attacks.

The report suggests that the crypto industry is worthy of vigilance because a large portion of its operations rely on open-source code, online infrastructure, and automated systems. Exchanges, wallets, smart contracts, and DeFi applications are constantly exposed to the internet environment, which inherently exposes them to ongoing security threats.

If stronger AI proxies can scan for contract vulnerabilities more quickly, identify misconfigurations, or assist in launching multi-step attacks, the response times of development teams and platforms could be further reduced. Compared to directly attacking the underlying blockchain, applications surrounding the encryption platform, the account system, and operational personnel may become more viable targets.

For the encryption industry, this means that security pressures may shift further from traditional hacking tools to AI systems that have the ability to execute autonomously. The platform will need to invest more in monitoring, early warning, and emergency response capabilities to deal with more frequent and automated attack methods.