MacMusic  |  PcMusic  |  440 Software  |  440 Forums  |  440TV  |  Zicos
models
Recherche

MINJA sneak attack poisons AI models for other chatbot users

jeudi 13 mars 2025, 10:04 , par TheRegister
Nothing like an OpenAI-powered agent leaking data or getting confused over what someone else whispered to it
AI models with memory aim to enhance user interactions by recalling past engagements. However, this feature opens the door to manipulation.…
Lire la suite sur TheRegister
https://go.theregister.com/feed/www.theregister.com/2025/03/11/minja_attack_poisons_ai_model_memory/

Voir aussi

models Google admits accidentally deleting Timeline data of Google Maps users due to a ‘technical issue’ BetaNews24 Mar
other New iOS Update Re-Enables Apple Intelligence For Users Who Had Turned It Off Slashdot22 Mar
minja Microsoft Tells Windows 10 Users To Buy New PCs Slashdot21 Mar
users Anthropic ajoute la recherche web à son chatbot Claude ZDNet.fr21 Mar
chatbot Gmail search gets worse as Google forces AI powered results on users BetaNews21 Mar
poisons FOSS infrastructure is under attack by AI companies OS News20 Mar
sneak Big Red, Microsoft roll out Azure database services for more mainstream Oracle users TheRegister20 Mar
attack GitHub suffers a cascading supply chain attack compromising CI/CD secrets InfoWorld19 Mar
models Meta's Llama AI Models Hit 1 Billion Downloads, Zuckerberg Says Slashdot18 Mar
other Roku's tests autoplaying home screen ads, and users feel tested BoingBoing18 Mar
minja Google acquisition target Wiz links fresh supply chain attack to 23K pwned GitHub repos TheRegister18 Mar
users Windows 11 update accidentally removes Copilot, causes issues for Citrix users ComputerWorld18 Mar
chatbot TikTok launches Security Checkup tool to help users keep their accounts protected BetaNews18 Mar
poisons Google Is Switching Legacy G Suite Users To Pooled Workspace Storage Slashdot18 Mar
sneak Microsoft accidentally cares about its users, releases update that unintentionally deletes Copilot from Windows OS News17 Mar
attack GitHub supply chain attack spills secrets from 23,000 projects TheRegister17 Mar
models Under Trump, AI Scientists Are Told to Remove ‘Ideological Bias’ From Powerful Models Wired: Tech.15 Mar
other Elon Musk's own AI chatbot — Grok — says U.S. better off with Democrats BoingBoing14 Mar
minja At last! New Hasleo Backup Suite Free gives Windows users the protection their data -- and system -- needs BetaNews14 Mar
users What is Llama? Meta AI’s family of large language models explained InfoWorld14 Mar
News copyright owned by their original publishers | Copyright © 2004 - 2025 Zicos / 440Network
Date Actuelle
mar. 15 avril - 21:30 CEST