Experimental News Clipping Site

Tag: sensitive applications

  • Cloud Blog: Chrome Enterprise improves work and personal profile separation and brings more advanced protections for unmanaged devices

    —

    by

    Kurt Seifried
    in

    Source URL: https://cloud.google.com/blog/products/chrome-enterprise/chrome-enterprise-improves-work-and-personal-profile-separation-and-brings-more-advanced-protections-for-unmanaged-devices/ Source: Cloud Blog Title: Chrome Enterprise improves work and personal profile separation and brings more advanced protections for unmanaged devices Feedly Summary: For many employees, the browser has become where they spend the majority of their working day. As more work is being done on the web, IT and security teams continue…

  • The Register: Does terrible code drive you mad? Wait until you see what it does to OpenAI’s GPT-4o

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.theregister.com/2025/02/27/llm_emergent_misalignment_study/ Source: The Register Title: Does terrible code drive you mad? Wait until you see what it does to OpenAI’s GPT-4o Feedly Summary: Model was fine-tuned to write vulnerable software – then suggested enslaving humanity Computer scientists have found that fine-tuning notionally safe large language models to do one thing badly can negatively…

  • Simon Willison’s Weblog: Aider Polyglot leaderboard results for Claude 3.7 Sonnet

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Feb/25/aider-polyglot-leaderboard/ Source: Simon Willison’s Weblog Title: Aider Polyglot leaderboard results for Claude 3.7 Sonnet Feedly Summary: Aider Polyglot leaderboard results for Claude 3.7 Sonnet Paul Gauthier’s Aider Polyglot benchmark is one of my favourite independent benchmarks for LLMs, partly because it focuses on code and partly because Paul is very responsive at evaluating…

  • Schneier on Security: Implementing Cryptography in AI Systems

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.schneier.com/blog/archives/2025/02/implementing-cryptography-in-ai-systems.html Source: Schneier on Security Title: Implementing Cryptography in AI Systems Feedly Summary: Interesting research: “How to Securely Implement Cryptography in Deep Neural Networks.” Abstract: The wide adoption of deep neural networks (DNNs) raises the question of how can we equip them with a desired cryptographic functionality (e.g, to decrypt an encrypted input,…

  • Simon Willison’s Weblog: Quoting Joanna Bryson

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Feb/20/joanna-bryson/ Source: Simon Willison’s Weblog Title: Quoting Joanna Bryson Feedly Summary: There are contexts in which it is immoral to use generative AI. For example, if you are a judge responsible for grounding a decision in law, you cannot rest that on an approximation of previous cases unknown to you. You want an…

  • Unit 42: Multiple Vulnerabilities Discovered in NVIDIA CUDA Toolkit

    —

    by

    Kurt Seifried
    in

    Source URL: https://unit42.paloaltonetworks.com/nvidia-cuda-toolkit-vulnerabilities/ Source: Unit 42 Title: Multiple Vulnerabilities Discovered in NVIDIA CUDA Toolkit Feedly Summary: Unit 42 researchers detail nine vulnerabilities discovered in NVIDIA’s CUDA-based toolkit. The affected utilities help analyze cubin (binary) files. The post Multiple Vulnerabilities Discovered in NVIDIA CUDA Toolkit appeared first on Unit 42. AI Summary and Description: Yes **Summary:**…

  • Slashdot: New Hack Uses Prompt Injection To Corrupt Gemini’s Long-Term Memory

    —

    by

    Kurt Seifried
    in

    Source URL: https://it.slashdot.org/story/25/02/12/0011205/new-hack-uses-prompt-injection-to-corrupt-geminis-long-term-memory?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: New Hack Uses Prompt Injection To Corrupt Gemini’s Long-Term Memory Feedly Summary: AI Summary and Description: Yes Summary: The text discusses a newly demonstrated attack by researcher Johann Rehberger that compromises Google’s Gemini chatbot by manipulating its long-term memory functionality through untrusted document summarization. The attack bypasses existing prompt…

  • The Register: Google torpedoes ‘no AI for weapons’ rules

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.theregister.com/2025/02/05/google_ai_principles_update/ Source: The Register Title: Google torpedoes ‘no AI for weapons’ rules Feedly Summary: Will now happily unleash the bots when ‘likely overall benefits substantially outweigh the foreseeable risks’ Google has published a new set of AI principles that don’t mention its previous pledge not to use the tech to develop weapons or…

  • Slashdot: OpenAI Teases ‘New Era’ of AI In US, Deepens Ties With Government

    —

    by

    Kurt Seifried
    in

    Source URL: https://yro.slashdot.org/story/25/01/30/2142256/openai-teases-new-era-of-ai-in-us-deepens-ties-with-government?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: OpenAI Teases ‘New Era’ of AI In US, Deepens Ties With Government Feedly Summary: AI Summary and Description: Yes Summary: OpenAI’s partnership with the US government, particularly with National Laboratories, aims to leverage AI for advancements in multiple fields, including national security, energy, and cybersecurity. This collaboration signifies a…

  • OpenAI : Trading inference-time compute for adversarial robustness

    —

    by

    Kurt Seifried
    in

    Source URL: https://openai.com/index/trading-inference-time-compute-for-adversarial-robustness Source: OpenAI Title: Trading inference-time compute for adversarial robustness Feedly Summary: Trading Inference-Time Compute for Adversarial Robustness AI Summary and Description: Yes Summary: The text explores the trade-offs between inference-time computing demands and adversarial robustness within AI systems, particularly relevant in the context of machine learning and AI security. This topic holds…