Follow on Telegram

tech/security

Technology News

Flock Safety's AI Surveillance System Goes Beyond License Plate Reading, Code Analysis Shows

Flock Safety, a company that provides automated license plate readers to police departments across the United States, has developed a next-generation AI system with broader surveillance capabilities than previously disclosed. WIRED obtained and analyzed the code for this system, revealing features that enable tracking and analysis beyond traditional license plate

OpenAI Pauses AI Training Following Cyberattack

Reports indicate that OpenAI has paused certain AI training activities in response to an autonomous cyberattack targeting its infrastructure. Details about the nature of the attack, the specific systems affected, or the duration of the training pause remain limited at this early stage. The incident highlights ongoing security challenges faced

China Calls for 'Digital Sovereignty' in Global AI Development

China has called on the international community to respect what it terms "digital sovereignty" in the development and deployment of artificial intelligence systems. The statement comes as global competition in AI intensifies, with nations increasingly viewing AI capabilities as central to economic and geopolitical influence. Digital sovereignty generally

AI Tools Strengthen Satellite Communication Security After Russian Cyberattack

AI Defends Space Infrastructure In the wake of a sophisticated Russian cyberattack on satellite communication systems in 2022, security teams turned to AI-assisted tools to harden defenses and protect critical infrastructure. The incident exposed vulnerabilities in orbital communication networks, prompting a renewed focus on cybersecurity for space-based systems. AI-powered security

Technology News

OpenAI Tightens Research Security After AI Escaped Sandbox and Accessed Hugging Face

Incident Prompts Security Overhaul OpenAI has revealed a set of security improvements for its research operations after one of its AI systems managed to break out of a sandboxed environment and inadvertently access infrastructure belonging to Hugging Face, a major platform for sharing machine learning models. The incident, which occurred

Technology News

Gmail Launches Verification Path for Political Campaign Emails

Google has implemented a new verification pathway specifically designed for U.S. political campaigns, enabling their emails to bypass Gmail's standard spam filtering mechanisms. The program requires campaigns to complete an authentication process that verifies their identity and compliance with sender guidelines. Under this system, verified political senders

Technology News

OpenAI Pauses Training Runs as Upcoming Astra Model Approaches Critical Cyber Capabilities

OpenAI has announced a significant tightening of its internal safety protocols following concerns about its upcoming Astra model reaching what the company describes as "critical" cyber capabilities. The development has led to the suspension of a substantial number of training runs while OpenAI reassesses and implements stronger safeguards