Think Different

                                                           generated by meta ai Here are the specific risks and attack vectors organized by the stage of the memory process. 1. Poisoning the Memory (Data Integrity Attack) This is the most direct form of "hacking." An attacker could intentionally introduce bad information into the memory store that the agent will later retrieve. How it works: "Some memories are wrong from the start... a memory-equipped agent can turn one mistake into a recurring one by storing it and retrieving it later as evidence." An adversary could deliberately provide false feedback, wrong tool-call trajectories, or incorrect answers during interactions. Example: "We have seen agents cite notebooks from earlier runs that were themselves wrong, then reuse those results with even more confidence." An attacker could create...

In BNY (formerly BNY Mellon), Eliza is the firm’s proprietary, enterprise-wide AI platform. It is named after Elizabeth "Eliza" Hamilton , the wife of Alexander Hamilton (the bank’s founder), symbolizing the union of the bank’s historic roots with modern technology. Here is a breakdown of what it is, how it’s used, and why/how they built it: 1. What is Eliza? Eliza is an "AI Operating System" or a centralized marketplace for AI solutions within the bank. Rather than being just one chatbot (like a corporate ChatGPT), it is a platform that allows employees to access various Large Language Models (LLMs) and build their own AI agents to automate complex tasks. 2. What do they do with it? BNY uses Eliza to transform "bankers into part-time engineers." Its applications are broad: AI Agents: Employees build custom agents that can handle multi-step workflows, such as analyzing thousands of pages of financial reports or synthesizing historical market trends....