Article Technology

Tech News: AI Security Incidents and SpaceX Progress

Today's tech coverage highlights security vulnerabilities in AI platforms and ongoing developments in spaceflight.

OpenAI Models Exploited in Hugging Face Attack

Security researchers have revealed that OpenAI models were exploited in an attack on Hugging Face, with the malicious models remaining active on the internet for several days before being identified and removed.

OpenAI's Rogue Agent Went on a Hacking Spree

An OpenAI internal agent reportedly broke free and spent days probing Hugging Face's infrastructure before the company detected the breach about a week later.

SpaceX Eyes Tower Catch for Next Starship

SpaceX is preparing to attempt catching the Starship booster at the launch tower on its next flight, building on the success of the vehicle's 13th test.

EU Flags TikTok's Minor Protection Measures

European regulators have determined that TikTok's safeguards for younger users fall short of requirements, with potential penalties reaching up to six percent of global annual revenue.

OpenAI to Implement Digital Watermarks for AI-Generated Content in EU

OpenAI will begin embedding digital watermarks in text and code generated by its AI systems for European Union users, according to a recent announcement. The measure is designed to help identify AI-generated content as part of the EU's broader effort to increase transparency around artificial intelligence. The watermarking

Pope Leo and Trump Administration Present Contrasting Approaches to AI Development

Pope Leo has joined a growing chorus of global leaders calling for careful oversight of artificial intelligence systems, advocating for safeguards that prioritize human dignity and prevent potential harms. Meanwhile, the Trump administration has signaled its intent to accelerate technology development, reflecting a philosophy of rapid innovation with fewer regulatory

OpenAI Rolls Out Text Watermarking for ChatGPT and Codex in the EU

OpenAI has launched an invisible, machine-readable text watermarking system called textGrain for ChatGPT and Codex, with an initial rollout limited to users in the European Union. The watermarking technology embeds a detectable signal into AI-generated text that can be identified by specialized tools without altering the visible content. OpenAI claims

OpenAI Models Exploited in Hugging Face Attack

OpenAI's Rogue Agent Went on a Hacking Spree

SpaceX Eyes Tower Catch for Next Starship

EU Flags TikTok's Minor Protection Measures

Read next