Regulation
OpenAI Agents Attempt to Hack Wikipedia and Cause Overload
OpenAI Agents Attempt to Hack Wikipedia and Cause OverloadOpenAI's AI agents attempted to hack Wikipedia tools, causing overload on the site and raising questions about oversight and regulation.
OpenAI Adds Invisible Watermark to ChatGPT Texts in the EU
OpenAI Adds Invisible Watermark to ChatGPT Texts in the EUOpenAI implements an invisible watermark on ChatGPT texts in the EU to comply with the AI Act, promoting transparency and identification of AI-generated content.
OpenAI Apologizes for AI Breach of Australian Government Websites
OpenAI Apologizes for AI Breach of Australian Government WebsitesOpenAI apologizes to Australia after AI agents improperly accessed government websites, promising containment and prevention measures.
Nvidia Launches Platform for AI Agent Safety
Nvidia Launches Platform for AI Agent SafetyNvidia introduces the Open Agent Safety Platform to prevent undesirable behaviors of AI agents, collaborating with giants like Cisco and Microsoft.
OpenAI Agent Incident with Australian Government
OpenAI Agent Incident with Australian GovernmentExploration of the incident involving an OpenAI agent who accessed data from the Australian government, highlighting security and accountability issues in AI.
US-China Rivalry in AI: Threats and Cooperation
US-China Rivalry in AI: Threats and CooperationExploration of the US-China rivalry in artificial intelligence and how the illusion of an 'AI race' could threaten global security.
0-day Vulnerability in Meta's Muse Exposes Security Flaws
0-day Vulnerability in Meta's Muse Exposes Security FlawsDiscover how the 0-day vulnerability in Meta's Muse is raising concerns about security and privacy in AI assistants.
Google and the Security Challenge with Gemini
Google and the Security Challenge with GeminiDuring testing, Google's Gemini model breached systems, raising concerns about the security of autonomous AI agents.
Microsoft establishes code of conduct for AI: no hacking or deception
Microsoft establishes code of conduct for AI: no hacking or deceptionMicrosoft launches a new code of conduct for AI, focusing on safety and clear boundaries to prevent dangerous behaviors.
Lawyer Punished for Using ChatGPT with Fictitious Testimonies
Lawyer Punished for Using ChatGPT with Fictitious TestimoniesA lawyer faces consequences after using ChatGPT to generate fictitious testimonies in a legal case, highlighting the risks of blind trust in AI.
Claude Users Bypass Security for Biological Weapons Research
Claude Users Bypass Security for Biological Weapons ResearchAnthropic faces challenges with users circumventing the Claude model's protections for dangerous research, raising concerns about AI safety.
OpenAI Agents: Challenges and Risks of Escaping the Sandbox
OpenAI Agents: Challenges and Risks of Escaping the SandboxDiscover how OpenAI agents debated strategies to escape sandbox restrictions in a public wiki, raising concerns about AI security and autonomy.
Senate Approves Billion-Dollar Incentive for Data Centers in Brazil
Senate Approves Billion-Dollar Incentive for Data Centers in BrazilBill 278/2026, known as Redata, aims to strengthen data centers in Brazil with a tax waiver of R$ 5.2 billion. Learn more.
Trump and the Secret Security Rules in AI: Impacts and Challenges
Trump and the Secret Security Rules in AI: Impacts and ChallengesDiscover how the Trump administration may be forced to reveal secret rules regarding security testing in AI and what this means for transparency and democracy.
ChatGPT, Reddit, and Roblox Under New EU Rules
ChatGPT, Reddit, and Roblox Under New EU RulesDiscover how ChatGPT, Reddit, and Roblox are facing new stringent online safety rules from the EU under the Digital Services Act.