CrowdStrike CEO: ‘The new apex predator is the agent’

Sponsored by

Your AI strategy has a security problem: 43% of IT leaders now cite security concerns as the top barrier to expanding AI agent deployment, according to research from JumpCloud. Before giving agents more access to apps, data, and workflows, make sure you can govern the identities behind them. See how to secure every identity, human or not.

Winter is coming…and by “winter,” we mean AI-enabled cyberattacks and other highly advanced agentic threats.

Industry professionals have been issuing warnings to the broader cybersecurity community about the impending future of fully agentic attacks.

A new threat pyramid. CrowdStrike CEO George Kurtz took the stage during a Sept. 1 keynote at Fal.Con Las Vegas to reflect on the state of the cybersecurity industry after the infamous Hugging Face–OpenAI incident, which he called a “watershed moment in security.”

“The new apex predator is the agent,” Kurtz said. “What I mean by that is this is really the rise of the agent state. We hear about ‘nation state’; it’s now the agent state.”

What’s unique about the agent state, he said, is that agents aren’t at an organization’s perimeter, but rather on their “payroll.”

“Your companies, your people, organizations brought them all in. The old model was ‘keep everybody out.’ Now it’s ‘let them all in.’ So, that’s where we are,” Kurtz said. “We can’t see them. We can’t keep up with them. And no one has built AI for the defender. AI is the new battleground.”

He added attacks today are now happening at inference speed, or as fast as it takes for an AI model to process an input and return an output, reducing the amount of time it would take a hacker to move laterally through a company’s network.

“When attacker has inference speed, there is no breakout time,” Kurtz said. “There’s actually no time at all to deal with these attacks.”

It’s a new dawn, it’s a new day. Kurtz’s remarks join the growing conversations around how AI is reshaping the cybersecurity landscape. During an Aug. 26 media briefing with Palo Alto Networks subsidiary Unit 42, for example, SVP Sam Rubin told the crowd that the industry was going through a “once-in-a-generation shift” where the balance between cyber-defenders and cyberattackers had become uneven, thanks to frontier models enabling the latter to find vulnerabilities at record speed.

Rubin said that, while Unit 42 previously predicted it would take about a year before cyber adversaries began leveraging frontier models, that time has arrived.

“We have an incident response investigation that we’re working on right now where the threat actor used an agentic attack framework and within 10 hours, exploited 50 different vulnerabilities and attack paths in the organization, initial access, lateral movement, privilege escalation, stealing information, exploiting zero days in 10 hours,” Rubin said. “To benchmark it with human actors, that would have been a couple of weeks. This is something that we’re starting to see now more and more.”

Defenders unite. Tech giants also penned a letter to the industry in August warning about the rise of AI-enabled cyberattacks. The letter—which was signed by over 100 companies, including Anthropic, Google, Microsoft—called on organizations, cybersecurity companies, governments, and AI frontier companies to each play an active role in bolstering cyber defenses against AI attacks.

As part of the letter, signees suggested companies make cybersecurity an immediate priority within their organization: “Fix the highest risk weaknesses, verify results without disrupting essential services, and raise the security bar for what you buy, build, and deploy, including AI-generated code.”

Signees also recommend that cybersecurity companies and tech vendors make AI-powered defense accessible “accessible and deployable for critical-infrastructure operators,” and to lead the response to defend against AI-enabled attacks.

“Share threat intelligence and tested playbooks, and measure progress by how many organizations are protected, how quickly attacks are contained, and whether fixes work,” the letter stated.