Cyber Security News

Hacking Editorial Brief: May 12, 2026

Google Confirms First AI-Assisted Zero-Day Discovery

Google's Threat Intelligence Group disclosed Monday that threat actors successfully used AI large language models to identify a previously unknown security vulnerability, marking a significant milestone in offensive security capabilities. The incident, which Google ultimately disrupted, represents the first confirmed case of criminal hackers weaponizing AI for autonomous vulnerability discovery rather than just exploit refinement. According to Google analysts, the attackers leveraged AI to accelerate reconnaissance and flaw identification before the company's defenses detected and blocked the exploitation attempt. The development confirms long-standing industry concerns that AI would lower barriers to zero-day discovery, with criminal groups now demonstrating capabilities previously associated with well-resourced nation-state actors.

Instructure Pays Ransom After Canvas Breach Disrupts Education

Instructure confirmed it paid an undisclosed ransom to cybercriminals following a breach of its Canvas learning management system that disrupted access for students during final examinations at multiple U.S. educational institutions. The attack, attributed to the ShinyHunters threat group, involved compromise of specific instructor account types and resulted in theft of personal data. Instructure stated the threat actors returned stolen data as part of the settlement, though the incident forced some universities to delay exams amid platform outages. The breach represents the second significant security incident for Instructure, highlighting persistent targeting of educational technology infrastructure containing sensitive student and faculty information.

Sources: The Information · Dark Reading · Inside Higher Ed · Reuters