Ethical considerations and safety measures are becoming increasingly important as AI models become more powerful. This includes addressing biases, preventing misuse, and ensuring responsible development.
A new study explores knowledge distillation to improve the safety of large language models (LLMs) in multilingual contexts. Results show that fine-tuning on "safe" data can paradoxically increase model vulnerability to jailbreak attacks, highlighting...
Anthropic, with its Claude model, appears to be shifting the focus in the AI race. The company is now focusing on aspects such as ethics and responsibility, in addition to pure computing power and model scalability.
A woman was victimized by AI-generated images. Strangers created nude images from her profile and opened an OnlyFans account in her name. The incident occurred during a surge in the generation of sexual images via AI, raising questions about the misu...
State-sponsored hackers are exploiting AI models like Gemini to refine phishing attacks and develop malware. Groups from Iran, North Korea, China, and Russia are leveraging AI for reconnaissance, social engineering, and malware development, increasin...
OpenAI has disbanded its mission alignment team, which focused on developing 'safe' and 'trustworthy' artificial intelligence. The team's leader will become OpenAI's Chief Futurist, with other members reassigned within the company.
Zoë Hitzig, an economist and researcher, resigned from OpenAI due to disagreements over ChatGPT's advertising strategy. She fears that the use of personal data shared by users could lead to manipulation, repeating past mistakes. Hitzig criticizes Ope...
AI assistants equipped with autonomous action capabilities raise serious concerns about data security. The article examines the risks associated with tools like OpenClaw, which offer extensive customization options but expose users to potential promp...
An AI-enhanced sinus surgery system experienced a significant increase in malfunctions, rising from eight to one hundred incidents. The investigation raises concerns about the safety and reliability of integrating AI into delicate medical procedures.
A podcast analyzes Ring's new features and raises concerns about mass surveillance. It also discusses how Apple's Lockdown Mode prevented the FBI from accessing a Washington Post reporter's iPhone, highlighting the importance of device security.
An OpenAI executive, who opposed the introduction of an "adult mode" in the chatbot, has reportedly been fired following allegations of discrimination. The executive has denied the allegations.
An AI chatbot from the U.S. Department of Health and Human Services, promoted by Robert F. Kennedy Jr., has generated questionable responses, suggesting foods suitable for rectal insertion and identifying the liver as the most nutritious human body p...
Researchers warn: a zero-click prompt injection vulnerability can leak data when AI agents meet messaging apps. An attacker can trick an AI agent into generating a data-leaking URL, which link previews may fetch automatically, exposing sensitive info...
Microsoft Azure researchers discovered that a single, unlabeled training prompt can disable the safety mechanisms built into several large language models (LLMs). The finding raises concerns about the robustness of current safeguards.
Healthcare researchers have found that AI chatbots could put patients at risk by giving shoddy medical advice. The quality of responses is compromised by users' failure to provide accurate details.
A user reports a severe prompt injection vulnerability in a self-hosted LLM system. During testing, a malicious prompt exposed the entire system prompt, highlighting the lack of adequate defenses against this type of attack. Traditional Web Applicati...
Anthropic invested millions of dollars in Super Bowl commercials to highlight its strategy, which rejects the insertion of advertising in chatbots, in contrast to other companies in the sector. The campaign aims to highlight a different approach to t...
Higgsfield.ai, a startup offering AI video creation tools, has generated outrage by claiming it contributed to artists' unemployment. The marketing stunt sparked a heated debate about the impact of AI on the creative job market.
The Department of Homeland Security’s (DHS) Inspector General has launched an investigation into Immigration and Customs Enforcement (ICE) regarding potential privacy abuses related to surveillance and biometric data programs. The investigation aims ...
