Episode #462 Intelligence artificielle et cyber sécurité : Interview d’Elie Bursztein Références : https://elie.net/talk/lessons-learned-from-developing-secure-ai-workflows-at-google-rsa-2024https://elie.net/talk/how-large-language-models-are-reshaping-the-cybersecurity-landscape-rsa-2024 https://elie.net/blog/hearthstone/i-am-a-legend-hacking-hearthstone-with-machine-learning-defcon-talk-wrap-up Quelques exemples de LLM malveillants : WolfGPT, WormGPT, EvilGPT, FraudGPT, PoisonGPT…https://www.sec.gov/news/press-release/2024-36https://c2pa.org/https://www.phonandroid.com/amazon-lutte-contre-les-livres-ecrits-par-lia-en-limitant-les-auteurs-a-3-publications-par-jour.htmlhttps://en.wikipedia.org/wiki/Prompt_engineering#Retrieval-augmented_generationhttps://www.vicarius.io/vsociety/posts/the-story-of-shadowray-cve-2023-48022https://bughunters.google.com/blog/5160301538967552/tensorflow-threat-model-and-security-guidelines-updatehttps://www.telegraph.co.uk/news/2018/03/17/calls-paris-sex-doll-brothel-closed-rape-fantasies/https://www.frontiermodelforum.org/https://owasp.org/www-project-top-10-for-large-language-model-applications/https://atlas.mitre.org/https://www.iso.org/standard/56581.html ISO 27090https://www.iso.org/standard/56582.html ISO 27091.2https://security.googleblog.com/2024/01/scaling-security-with-ai-from-detection.htmlhttps://elie.net/blog/ai/on-the-consequences-of-ai-workforce-entering-the-markethttps://everydayrobots.com/https://twitter.com/mckaywrigley/status/1762175776726663267https://www.reddit.com/r/singularity/comments/1bkcjs4/gemini_15_makes_a_scholarly_connection_that_took/
The post Interview d’Elie Bursztein au sujet de l’IA appeared first on NoLimitSecu.
