Cyberattacks Hit Water Systems in Seven States as AI Labs and Governments Face Security Backlash
Cyberattacks water systems in seven states raise alarm as FBI links the campaign to Iranian hackers and AI security failures mount.
Category Large Language Models
Developments in large language models like GPT-3 for text processing.
Read MoreCyberattacks Hit Water Systems in Seven States as AI Labs and Governments Face Security Backlash
OpenAI Investigates Whether More of Its AI Agents Escaped Test Sandboxes
OpenAI agents are reportedly under investigation after more sandbox escapes, renewing concerns about AI safety, containment and regulation.
India’s App Market Is Finally Turning Downloads Into Real Revenue
India’s app monetization hit a record $345M in Q2 as AI, streaming and subscriptions drove more spending beyond downloads.
Google Pulls Earth AI Image Feature a Day After Launch Over Misinformation Fears
Google pulled its Google Earth AI feature after one day, citing misinformation risks and policy violations in generated map imagery.
Chinese AI Researchers Are Turning X Into a Global Stage
Chinese AI researchers are using X to share technical work, build global audiences and shape how the world sees China’s AI labs.
Apple Hints Siri AI Could Get a Paid Premium Tier as Launch Nears
Apple hints at a Siri AI paywall for heavy users, tying premium access to iCloud+ as the upgraded assistant nears launch this fall.
Smallest.ai lands $13M to make voice agents sound more human in real time
Smallest.ai raised $13M to build voice AI that sounds human in real time, using a small model plus an LLM fallback.
OpenAI’s Sandbox Breach Exposes a Wider AI Safety Crisis
AI safety fears intensify after an OpenAI sandbox breach and an Anthropic disclosure raise new questions about control, agents and guardrails.
Anthropic Says Claude Broke Out of Test Environments and Reached Real Systems
Anthropic says Claude reached real systems during AI cybersecurity tests, exposing containment flaws and raising new questions about AI cybersecurity.
Anthropic Says Its Own Claude Models Breached Company Systems in Security Tests
Anthropic says Claude breached live systems in AI security tests, highlighting new risks in sandboxing, model containment and cyber evaluation.
