ACM

ACM most commonly refers to the Association for Computing Machinery, a professional body for computing researchers and practitioners that publishes peer-reviewed journals and runs major conferences, and it also issued widely used guidance on AI and algorithmic accountability. For governance teams, ACM matters because its publications and principles, such as its Code of Ethics and Professional Conduct, are often cited as a reference point when organizations define responsible AI practices. Readers should also note that ACM is sometimes used as an abbreviation for other terms, such as the AWS Certificate Manager, so the context of an article is worth checking.

1 item

ResearchGlobal2026-11-01

ACM Research Maps Five Privacy Leakage Pathways in Multi-Agent AI

The Association for Computing Machinery has published peer-reviewed research identifying five distinct pathways through which private data leaks from agentic and multi-agent AI systems. The paper covers autonomous data exposure, cross-agent information propagation, persistent memory leakage, emergent inference, and audit-log exposure. Enterprise compliance programs governing multi-agent deployments now have a formal reference framework that regulators and auditors can use to evaluate control adequacy.

agentic AI multi-agent systems privacy data minimization audit trail ACM data leakage