Experimental News Clipping Site

Tag: safe

  • Slashdot: OpenAI Releases First Open-Weight Models Since GPT-2

    —

    by

    Kurt Seifried
    in

    Source URL: https://slashdot.org/story/25/08/05/1848236/openai-releases-first-open-weight-models-since-gpt-2?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: OpenAI Releases First Open-Weight Models Since GPT-2 Feedly Summary: AI Summary and Description: Yes Summary: OpenAI’s release of two open-weight language models, gpt-oss-120b and gpt-oss-20b, marks a significant development in the AI landscape since 2019. These models enable local deployment on consumer devices and introduce advanced capabilities such as…

  • Simon Willison’s Weblog: Claude Opus 4.1

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Aug/5/claude-opus-41/ Source: Simon Willison’s Weblog Title: Claude Opus 4.1 Feedly Summary: Claude Opus 4.1 Surprise new model from Anthropic today – Claude Opus 4.1, which they describe as “a drop-in replacement for Opus 4". My favorite thing about this model is the version number – treating this as a .1 version increment looks…

  • New York Times – Artificial Intelligence : OpenAI to Open-Source Some of the A.I. Systems Behind ChatGPT

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.nytimes.com/2025/08/05/technology/openai-artificial-intelligence-chatgpt.html Source: New York Times – Artificial Intelligence Title: OpenAI to Open-Source Some of the A.I. Systems Behind ChatGPT Feedly Summary: In a major shift, the company is “open sourcing” two A.I. systems, freely sharing the technology with outside researchers and businesses. AI Summary and Description: Yes Summary: The text highlights a significant…

  • OpenAI : Estimating worst case frontier risks of open weight LLMs

    —

    by

    Kurt Seifried
    in

    Source URL: https://openai.com/index/estimating-worst-case-frontier-risks-of-open-weight-llms Source: OpenAI Title: Estimating worst case frontier risks of open weight LLMs Feedly Summary: In this paper, we study the worst-case frontier risks of releasing gpt-oss. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as capable as possible in two domains: biology and…

  • OpenAI : Open Weights and AI for All

    —

    by

    Kurt Seifried
    in

    Source URL: https://openai.com/global-affairs/open-weights-and-ai-for-all Source: OpenAI Title: Open Weights and AI for All Feedly Summary: AI’s next frontier isn’t just about capability—it’s about who gets to use it. Our mission to put AI in the hands of as many people as possible is what drives us. Today’s release of our most capable open-weights models is a…

  • Slashdot: AI Is Listening to Your Meetings. Watch What You Say.

    —

    by

    Kurt Seifried
    in

    Source URL: https://yro.slashdot.org/story/25/08/05/1546241/ai-is-listening-to-your-meetings-watch-what-you-say?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: AI Is Listening to Your Meetings. Watch What You Say. Feedly Summary: AI Summary and Description: Yes Summary: The text highlights a significant issue related to privacy breaches in AI meeting transcription software that automatically shares unintended personal conversations with all meeting participants. This raises concerns for security and…

  • Embrace The Red: Amp Code: Arbitrary Command Execution via Prompt Injection Fixed

    —

    by

    Kurt Seifried
    in

    Source URL: https://embracethered.com/blog/posts/2025/amp-agents-that-modify-system-configuration-and-escape/ Source: Embrace The Red Title: Amp Code: Arbitrary Command Execution via Prompt Injection Fixed Feedly Summary: Sandbox-escape-style attacks can happen when an AI is able to modify its own configuration settings, such as by writing to configuration files. That was the case with Amp, an agentic coding tool built by Sourcegraph. The…

  • The Register: Chained bugs in Nvidia’s Triton Inference Server lead to full system compromise

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.theregister.com/2025/08/05/nvidia_triton_bug_chain/ Source: The Register Title: Chained bugs in Nvidia’s Triton Inference Server lead to full system compromise Feedly Summary: Wiz Research details flaws in Python backend that expose AI models and enable remote code execution Security researchers have lifted the lid on a chain of high-severity vulnerabilities that could lead to remote code…

  • Slashdot: An Illinois Bill Banning AI Therapy Has Been Signed Into Law

    —

    by

    Kurt Seifried
    in

    Source URL: https://slashdot.org/story/25/08/05/148238/an-illinois-bill-banning-ai-therapy-has-been-signed-into-law?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: An Illinois Bill Banning AI Therapy Has Been Signed Into Law Feedly Summary: AI Summary and Description: Yes Summary: Illinois has enacted legislation that prohibits AI from serving as an independent therapist and establishes strict guidelines for using AI in mental health care. This law ensures that therapeutic services…