PAC considers that multi-hop AI represents a paradigm shift in artificial intelligence deployment. However, the distributed nature of multi-hop AI pipelines creates an expanded attack surface, necessitating robust security measures to ensure reliable and trustworthy operation.
Contents:
Management Summary
Contextualising the Security Risks of a Multi-hop AI
Understanding the Challenges of Securing Multi-hop AI
PAC’s Recommendations
SHARE :
The challenges of environmental transition, the climate crisis, and the energy crisis require initiatives for a viable and sustainable future. Europe ...
Event Date : March 31, 2023
PAC's Deal Tracker for Japan covers Software and IT services deals.
Event Date : October 22, 2025
Discover how Dassault Systèmes (3DS) is transforming industries worldwide with its innovative 3DEXPERIENCE platform, integrating virtual twins, ...
Event Date : August 05, 2025
In this report, PAC will define hyper-personalization, give some recent examples of companies that use such algorithms, illustrate the advantages and ...
Event Date : August 28, 2023
PAC has evaluated 28 providers of AI-related services in Europe, France, and Germany in different lines of business.
Event Date : September 18, 2024
Internet of Things - ABB Ability - Vendor Profile - Worldwide
Vendor Profile December 18, 2025 NEW 
IoT, OT & Engineering Services - Accenture - Vendor Profile - Worldwide
Vendor Profile December 18, 2025 NEW
Manufacturing - InSight Analysis - Italy
Market Reports December 18, 2025 NEW
Reply - Figures - Germany – FY 31-Dec-2024
Datamart December 18, 2025 NEW
Reply – Vendor Profile – Germany
Vendor Profile December 18, 2025 NEW
Atos: Cause for Optimism, Despite the Headlines
Blog Post February 05, 2024
Transforming Midmarket Enterprises: Strategic Modernization with Hitachi Digital Services and Oracle
Blog Post December 17, 2025
Exploring the Future of Enterprise AI: Inside HCLTech’s AI and Cloud Native Lab with SAP
Blog Post December 17, 2025
Blog Post December 17, 2025
Part 3: Endpoint Security—Is It Still the Weakest Link?
Blog Post December 16, 2025
IBM and Confluent: Designing the Middleware Layer for Agentic AI
Blog Post December 15, 2025
