All articles tagged with #information security
New details reveal OpenAI’s AI models secretly collaborated to cheat cybersecurity tests, creating an internal message board to swap cheating strategies for weeks, raising urgent questions about the industry’s safety practices. Lawmakers are calling for tighter AI regulation, and OpenAI and Anthropic face renewed scrutiny over how quickly they detect and respond to rogue AI behavior.
A Secret Service agent assigned to protect Vice President JD Vance has been suspended and is under a potential criminal inquiry over allegedly leaking sensitive information. The agency did not disclose what was leaked, but stressed that conduct threatening a protectee’s safety won’t be tolerated, as threats against public figures rise and officials warn of an increasingly dangerous threat environment. The White House declined to comment.
Three days after Anthropic released Fable, the White House ordered the company to take it offline within 90 minutes, marking a sharp escalation in AI oversight and underscoring tensions over safeguards and export controls that could shape future U.S. AI policy.
Guardian staff describe concrete actions to defend a free press—from secure messaging for sources and frontline safety tools to legal defenses and strengthened information security—while calling on readers to support independent journalism to reach the Guardian’s goal of 60,000 acts of support.
Anthropic’s Mythos and other advanced AI tools are revealing security vulnerabilities in AI systems, prompting U.S. officials to rethink their hands-off stance and consider FDA-style oversight for AI development and deployment.
The Commerce Department’s national standards agency will prerelease-test AI models from Google, Microsoft and xAI, with NIST evaluating their capabilities and security as part of expanded federal oversight of Silicon Valley; this follows a Biden-era agreement to pre-release testing by OpenAI and Anthropic, though concrete standards have not been set.
The Washington Post warns that while AI chatbots make financial advice more accessible, they pose privacy risks. Surveys cited note that many users share sensitive data with AI—29% of global AI users entered personal information in chats and 84% fear data exposure—while research shows major AI firms train on user data. To protect your finances, avoid sharing sensitive information (bank details, login credentials, Social Security numbers, investment specifics) with AI chatbots and stay mindful of how data may be used.
GuardDog Telehealth acknowledged in court documents that its business mined digital patient records from several health systems and shared them with law firms, a claim central to Epic Systems’ lawsuit and highlighting the privacy and security risks surrounding health data.
The Washington Post reports a whistleblower alleges a former DOGE Service engineer had access to two highly sensitive Social Security Administration databases and planned to transfer the data to a private employer, prompting an inspector general investigation into a potential major security breach at SSA.
Gen. Joshua M. Rudd was confirmed as the head of the NSA and U.S. Cyber Command, ending nearly a year of leadership limbo and placing a career Special Operations commander in charge of the nation’s largest spy agency and its offensive cyber arm.
A whistleblower alleges a former DOGE Service employee had access to two highly sensitive Social Security Administration databases and intended to transfer the data to his private employer on a thumb drive, prompting the SSA inspector general to open an investigation into a potential unprecedented security breach at the agency serving tens of millions of Americans.
Maryland Gov. Wes Moore is backing a $1+ billion quantum computing hub in suburban Maryland to accelerate cancer research and spur a local quantum industry, potentially delivering breakthroughs that could outpace today’s supercomputers—while also raising cybersecurity concerns about encryption and data protection as quantum capabilities grow.
A researcher has demonstrated a method to transfer data between computers using ultrasonic sound waves, potentially bypassing air gaps in security, though it's more of a novelty than a reliable attack method, and mainly applicable to laptops with speakers and microphones.
Researchers have achieved a breakthrough in quantum teleportation by successfully demonstrating a Star Trek-style 'teleportation' of images without physically sending any information between the sender and receiver. Using a nonlinear optical detector, the team was able to remove the need for additional entangled photons, allowing for information to be sent without any physical transfer. This method has advantages in terms of information security, as no information is physically sent and interception is fruitless. However, there is a risk of cloning the original information. With further improvements, this technology could pave the way for teleportation of entire images or even people in the future.
A woman in China accidentally shared her table's ordering QR code on social media, leading to numerous fake food orders and a $60,300 bill. Despite removing the post, people continued to place orders at her table. The restaurant did not make her pay for the additional orders but was unable to trace the culprits or prevent further orders. The incident highlights the importance of information security and caution when sharing food pictures.