PAC views the European Union Artificial Intelligence (EU AI) Act, required to be implemented and enforced by 2027, as a pivotal development in AI governance, introducing a risk-based regulatory framework that strikes a balance between innovation, accountability, and security. The Act categorises AI capabilities by potential harm, imposing stricter obligations on high-risk capabilities such as generative and agentic AI. It expands cybersecurity requirements to include algorithmic integrity, data provenance, and decision explainability, mandating secure, ethical, and transparent AI operations.
Compliance requires a comprehensive organisational response that integrates governance, operational processes, and cultural change. PAC considers that organisations must adopt secure development practices, maintain traceable documentation, and embed continuous risk management across the AI lifecycle. Collaboration between cybersecurity, data science, and compliance teams is essential to ensure human oversight and verifiable AI outputs. Accountability also extends to third-party suppliers, necessitating enhanced due diligence and monitoring.
PAC highlights that beyond regulatory alignment, the Act offers strategic advantages because organisations that embed its principles can build trust, improve AI reliability, and mitigate reputational and operational risks. Compliance promotes effective data governance and cross-functional collaboration, thereby supporting innovation and growth. Ultimately, a cybersecurity-first approach to AI enables scalable, responsible deployment, aligning ethical governance with strategic objectives and enhancing competitiveness in an AI-driven landscape.
Recommended advisory: PAC Leadership Session – Cybersecurity Compliance
SHARE :
This document provides market volumes, growth rates and forecasts for the IT security market in Slovakia for the 2024-2030 period.
Event Date : March 24, 2026
This Excel document is part of the company profiles PAC publishes every year at local, regional and worldwide level.
Event Date : September 26, 2025
This Excel document is part of the company profiles PAC publishes every year at local, regional and worldwide level.
Event Date : August 25, 2025
LTIMindtree (LTIM) is a US-centric IT services vendor (75%) with a client base organized around three verticals that represent ~80% of revenues: BFSI ...
Event Date : September 04, 2025
The role of contact centers has increased in prominence with the advent of the digital age and the impact of the recent global pandemic. However, ...
Event Date : April 27, 2023
Public Sector - Market View - Netherlands
Market Reports October 22, 2026 NEW 
T-Systems - Figures - Netherlands - FY 31-Dec-2025
Datamart October 21, 2026 NEW
T-Systems - Vendor View - Netherlands
Vendor Profile October 21, 2026 NEW
Atos - Figures - Belgium - FY 31-Dec-2025
Datamart October 20, 2026 NEW
Vendor Profile October 20, 2026 NEW
Atos: Cause for Optimism, Despite the Headlines
Blog Post February 05, 2024
Re-Imagine Enterprise-Scale Cyber Defence Through Agentic AI Platforms
Whitepaper & Trend Studies October 20, 2026
Key Takeaways from Hitachi Hour Krakow
Blog Post October 06, 2026
AI Drives CONTACT Software’s Industrial Expansion
Blog Post September 28, 2026
Sovereign Cortex in Germany: Beyond Data Residency
Blog Post September 25, 2026
Cloud X Summit by STACKIT: "We have come to stay"
Blog Post September 23, 2026
