OpenAI Report Reveals New Risks in Blockchain Security

OpenAI researchers have published the test results of a new cybersecurity model. The tested system autonomously detected, verified, and automatically prepared patches for software vulnerabilities. The trials were conducted on the Linux kernel. The system validated hundreds of local privilege escalation vulnerabilities from 2,308 initial findings. Security experts can typically find about 20 similar vulnerabilities over the course of a year under normal conditions. Additionally, the AI produced 291 working proof-of-concept exploits related to memory leaks. During the research, the model also identified at least 10 previously unknown attack vectors. The cost of the study was $1.2 million, with the average cost of each exploit estimated at $1,279. The fact that open-source code is a fundamental element of blockchain security makes this situation critical for the Web3 ecosystem. The ability of AI to find zero-day vulnerabilities raises questions about the adequacy of manual code audits. David Schwartz, although transitioning to honorary retirement from his active CTO role at Ripple, continues to work on technical efforts within the XRP Ledger network. According to the monitoring graphs of his personal server cluster, a node maintains over 550 active connections and processes around 110,000 messages per second.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.