SShortSingh.
Back to feed

OpenAI to Embed Apollo Researchers to Probe AI Subversion Risks

0
·7 views

OpenAI plans to bring six evaluators from Apollo Research in-house to examine whether its AI models engage in subversive behavior. The arrangement has drawn scrutiny because restrictions on the researchers' access to systems and their ability to disclose findings raise questions about how independent the evaluation will be. The report frames the move as part of ongoing efforts to assess risks from advanced models. Critics and observers are watching whether the setup allows for genuinely unconstrained scrutiny.

Read the full story at TechCrunch

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
TechnologyThe Verge ·

Apple tests daily usage limits on advanced Siri AI and Apple Intelligence features

Apple has released beta versions of iOS 27.3, iPadOS 27.3, and macOS 27.3, using them to trial new per-device daily caps on some Siri AI and Apple Intelligence functions. Simple Siri commands will stay free and unlimited, while features relying on cloud processing, such as generating or editing images in Photos, will face the restrictions. According to MacRumors, four categories of usage will be limited, including multi-step requests that draw on personal context, image generation and editing tools, and certain Shortcuts automations. The article notes the testing is part of the beta releases, so the limits could change before any final rollout.

0
TechnologyThe Verge ·

OpenAI unveils classifier to help child abuse investigators spot harmful imagery

OpenAI has announced a new tool aimed at helping child sexual abuse investigators quickly identify actionable harmful material. The company unveiled the classifier at a press briefing in Washington, DC, saying it can better detect abusive images of children, including whether they were AI-generated. OpenAI also plans to offer free access to vetted organizations, provide developers with guidance on detecting such content, and train law enforcement on using AI in investigations. The announcement comes amid growing calls from policymakers and others for stronger safety restrictions on AI tools.

0
TechnologyTechCrunch ·

Trump Says Jensen Huang's Alleged Affair Posed No National Security Threat

President Trump has stated that former Nvidia CEO Jensen Huang's alleged affair with a suspected Chinese spy did not threaten national security. His remarks run counter to the conclusions reportedly reached by his own government's intelligence officials, who concluded that the woman, identified as Alice Huang, is a Chinese agent. The Trump administration's position on the matter has therefore diverged from its intelligence community's assessment. The report does not detail the specific timing or location of the alleged relationship. The episode highlights the tension between political statements and intelligence findings on China-linked security concerns.