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

MINJA sneak attack poisons AI models for other chatbot users

Thursday March 13, 2025. 10:04 AM , from 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.…
Read more at TheRegister
https://go.theregister.com/feed/www.theregister.com/2025/03/11/minja_attack_poisons_ai_model_memory/

Related News

models At last! New Hasleo Backup Suite Free gives Windows users the protection their data -- and system -- needs BetaNewsMar 14
other What is Llama? Meta AI’s family of large language models explained InfoWorldMar 14
minja India investigates whether Uber makes iPhone users pay more to ride TheRegisterMar 14
users Mozilla Warns Users To Update Firefox Before Certificate Expires SlashdotMar 14
chatbot ViewSonic ColorPro VP2788-5K is the 5K monitor Mac and Windows users have been waiting for BetaNewsMar 13
poisons DeepSeek outperforms US models in new AI Trust Score BetaNewsMar 13
sneak How DeepSeek innovated large language models InfoWorldMar 13
attack AI models hallucinate, and doctors are OK with that TheRegisterMar 13
models Google unveils Gemma 3 multi-modal AI models InfoWorldMar 13
other Apple has locked me in the same monopolistic cage Microsoft's built for Windows 10 users TheRegisterMar 12
minja Official HP toner not official enough after dodgy update, say users TheRegisterMar 11
users DOGE deploys AI chatbot at GSA amid workforce reductions ComputerWorldMar 11
chatbot Eight days later, Microsoft Outlook users still struggle on iOS devices TheRegisterMar 11
poisons AI Gives Users ‘Superpowers of Cognitive Mobility,’ Says LinkedIn Cofounder eWeekMar 10
sneak Microsoft is developing its own AI models to compete with OpenAI ComputerWorldMar 10
attack Being high on MDMA during Hamas attack on music festival may have protected victims from trauma BoingBoingMar 10
models Microsoft Reportedly Develops LLM Series That Can Rival OpenAI, Anthropic Models SlashdotMar 8
other DOGE Has Deployed Its GSAi Custom Chatbot for 1,500 Federal Workers Wired: Tech.Mar 8
minja Microsoft reportedly struggling to build its own reasoning models to rival OpenAI InfoWorldMar 7
users 3.6 Million Lies: Is Your AI Chatbot Spewing Russian Disinformation? eWeekMar 7
News copyright owned by their original publishers | Copyright © 2004 - 2025 Zicos / 440Network
Current Date
Mar, Fri 14 - 16:21 CET