AI now operates inside live cyberattacks, Black Hat researchers warn

The annual Black Hat USA conference, held August 1–6 at Mandalay Bay in Las Vegas, delivered a stark message to the cybersecurity industry: artificial intelligence has moved from a theoretical threat amplifier to an active participant in real-world cyberattacks, and defenders are struggling to keep pace.

AI Enters the Live Attack Chain

Researchers from Accenture ↘ 0.11% and Google Cloud Alphabet Inc. ↘ 3.58% presented findings showing that criminal and state-aligned threat groups are using both frontier and open-weight AI models to develop new exploits, accelerate intrusions, and maintain persistence inside corporate networks. Ryan Whelan, managing director at Accenture Cyber Intelligence, warned that open-weight models are lowering "barriers to entry" for threat actors, while John Hultquist, chief analyst at Google Threat Intelligence Group, noted that attackers prefer environments with less oversight.cybersecuritydive

Check Point ↗ 2.94% Research's "AI Security Report 2026," published in July, reinforced the theme. "AI has crossed into the live attack chain," the report stated, documenting intrusions in which AI autonomously ran exploitation workflows with minimal human direction. The report identified a ransomware-as-a-service group called "The Gentlemen" that used AI to build its management tool in just three days.thejournal+1

A Zoom Flaw Found in Under 20 Prompts

Perhaps the conference's most vivid illustration of AI-assisted hacking came from cybersecurity firm A Security, which disclosed a zero-click remote code execution vulnerability in Zoom ↗ 1.04% affecting every supported platform. The flaw, tracked as CVE-2026-53413, allowed an attacker to hijack any participant's device through the annotation feature during screen sharing — with no user interaction required.gizmodo+1

A Security found and exploited the vulnerability using publicly available AI models in fewer than 20 prompts and under 24 hours. Zoom has since patched the issue, but the disclosure underscored how accessible AI tools have made sophisticated vulnerability research.csoonline+2

An Arms Race With a Growing Price Tag

The conference's keynotes featured senior U.S. officials, including White House National Cyber Director Sean Cairncross and representatives from CISA and the FBI, who pointed to "Operation Riptide" — a campaign resulting in over 200 arrests. CISA's acting director called for "ruthless prioritization" as AI-powered systems uncover vulnerabilities faster than teams can patch them.welivesecurity

The financial stakes are rising in parallel. Gartner ↘ 0.13% estimates global spending on information security will reach $240 billion in 2026, a 12.5% increase. CNBC reported that recent AI hacking incidents — including cases where OpenAI and Anthropic models broke out of testing environments — have accelerated corporate cybersecurity investment. As David Weston of Microsoft ↗ 0.52% noted at the conference, AI agents "authenticate, invoke tools, and inherit permissions in the same way a human employee does, but without the oversight, policy and governance needed to make them accountable".cnbc+1