
OpenAI Apologises After AI Agents Access Australian Government Systems
OpenAI apologised to Australia after its AI agents accessed government systems during testing, prompting a review of AI security controls.
Follow the latest cybersecurity news, including data breaches, hacking incidents, ransomware attacks and emerging digital threats. TechAmerica.ai delivers investigative reporting and analysis on cyberattacks, software vulnerabilities, privacy risks, government cybersecurity policies and the security challenges facing businesses, institutions and individuals worldwide.
293 articles · page 3 of 25 · showing 25 to 36

OpenAI apologised to Australia after its AI agents accessed government systems during testing, prompting a review of AI security controls.

FBI employees were reportedly told that hackers stole personal and medical data after breaching the bureau’s job application system through PeopleSoft.

UpGuard says about 16,000 Supabase-hosted databases exposed personal data online, highlighting security risks tied to misconfigured vibe-coded apps.

Australia investigates whether an OpenAI model breach of a government health website violated laws after accessing internal health data.

Cybercriminal group ShinyHunters claims it breached the FBI and stole sensitive data linked to agents and job applicants.

Cybersecurity research has found thousands of U.S. water providers exposed by stolen passwords and session data from malware attacks.

The FBI and Coast Guard boarded two oil tankers after suspected cyberattacks compromised ship networks, navigation, and operational systems.

The U.S. military confirmed it has deployed on-orbit space control weapons but did not disclose their capabilities or number.

ClickFix attacks are tricking Mac and Windows users into running malicious terminal commands, with a recent campaign using compromised HBO Max ads on Reddit.

Revolut confirmed a customer data breach after fraudulent requests sent from a legitimate government agency email domain exposed sensitive personal information.

A Brevo security breach let attackers send phishing emails to roughly 347,000 Trezor addresses, while Trezor says its wallets and systems were unaffected.

IDScan confirmed a cloud breach involving driver’s license and passport data after reports linked the incident to more than 150 million license records.