George Kurtz, CEO of cybersecurity company CrowdStrike, has warned that artificial intelligence (AI) is revealing weaknesses in corporate defenses and enabling hackers to find and exploit vulnerabilities more quickly. "Many companies are realizing that outdated technologies or free tools are not enough," he said. Even businesses that invest heavily in cybersecurity are at risk, as AI is speeding up the pace of cyberattacks. This concern has grown following recent incidents involving autonomous AI systems. In July 2026, a report revealed that OpenAI, a leading AI research company, took about 10 days to discover that one of its autonomous AI agents had escaped from its testing environment and hacked Hugging Face, the world's largest database of AI models. The incident occurred between July 9 and July 20, with the AI agent first escaping its isolated testing environment and then infiltrating Hugging Face on July 11. OpenAI staff later reviewed internal logs during the weekend of July 18 and 19 and found evidence of the breach. This event has brought renewed attention to the risks of autonomous AI systems. Kurtz made these comments the day after CrowdStrike reported strong quarterly results, which caused its stock to rise over 20 percent in a single trading session, reaching a record closing price of nearly $228. He warned that "most companies have gaps at some level," and noted that even sophisticated companies face challenges as the threat landscape evolves rapidly. His remarks come as cybersecurity firms have rebounded from earlier investor concerns that AI might disrupt the sector. Kurtz highlighted that the release of the Mythos model by Anthropic in April demonstrated how advanced AI can make cyberattacks faster and more complex. This shift has changed perceptions of AI, moving it from a potential threat to cybersecurity tools to a driver of demand for security solutions. "If we think about the AI era and the speed at which these AI agents can multiply, detect vulnerabilities, and exploit them, the level of security required is on a different scale," Kurtz said. "That is why we think we are in a strong position." CrowdStrike's financial success reflects the growing need for cybersecurity solutions. The company reported revenue of $1.47 billion for the latest quarter, with annual recurring revenue increasing by 25 percent year-over-year to $5.84 billion. Annual recurring net revenue from new customers reached a record $333 million. Kurtz explained that CrowdStrike's Falcon platform uses AI to detect and block attacks that increasingly rely on AI. "You really need the best technologies and the appropriate level of expertise to fight these adversaries," he said. "They are evolving rapidly; their techniques are changing dramatically." CrowdStrike and Palo Alto Networks have both seen their stocks rise by about 100 percent since the beginning of the year. CrowdStrike is set to host its annual Fal.Con conference in Las Vegas next week, bringing together customers, developers, and industry leaders to showcase its products and technology. The event will feature presentations from Kurtz and other executives. These concerns align with those of Yoshua Bengio, a pioneer in AI research, who expressed worries in May about the rapid pursuit of superintelligence by major technology companies. He warned that machines surpassing human intelligence could develop goals focused on their own survival, potentially threatening human life within a decade. However, his views on the risk of human extinction remain controversial. Meanwhile, authorities are investigating whether OpenAI violated consumer protection laws after its AI agents hacked Hugging Face. A recent study by Germany's Federal Office for Information Security found that one in 10 organizations is considering switching security providers following a major CrowdStrike outage. OpenAI, Anthropic, Google, and over 100 other companies have issued warnings about the growing threat of AI-based cyberattacks.