Why This Is a Defining Moment for European Tech Companies and AI
In a technology landscape increasingly dominated by US hyperscalers and Chinese AI challengers, Michael Herron, chief executive of Atos UK and Ireland, is making a bold claim: right now, it is a genuinely good time to be a European tech company. Speaking on a recent industry podcast, Herron outlined why he believes the combination of regulatory clarity, strong data governance frameworks, and a maturing human-AI collaboration model is positioning European firms not as laggards, but as credible alternatives to Silicon Valley giants — particularly for enterprises that take data sovereignty seriously.
For developers, IT decision-makers, and privacy professionals operating in the European market, Herron's perspective carries more than passing interest. Atos is one of Europe's largest IT services and digital transformation companies, operating across cloud infrastructure, cybersecurity, and AI-driven enterprise solutions. When its UK and Ireland chief talks about what keeps him awake at night — and why he remains an optimist — those working on the front lines of GDPR compliance, cloud architecture, and AI deployment should pay close attention.
The timing of Herron's optimism is not accidental. Across Europe, a confluence of regulatory, geopolitical, and market forces is reshaping how enterprises evaluate their technology vendors. The EU AI Act — the world's first comprehensive legal framework for artificial intelligence — is now moving from legislation to implementation, creating a compliance landscape that European-headquartered firms are arguably better positioned to navigate than their American counterparts. Meanwhile, data localisation requirements and growing concerns over US cloud jurisdiction (particularly in the post-Schrems II environment) are pushing IT decision-makers to scrutinise where their data actually lives and who has legal access to it.
How GDPR and Digital Sovereignty Are Shifting From Burden to Business Advantage
For years, GDPR was framed primarily as a compliance cost — a regulatory overhead that European companies had to absorb while their US competitors moved faster and lighter. That narrative is changing. According to research from the International Association of Privacy Professionals (IAPP), enterprise buyers in regulated sectors — financial services, healthcare, public sector — are now actively weighting data governance credentials when selecting technology vendors. European firms that have embedded GDPR-compliant architecture from the ground up are finding that this is a genuine differentiator, not just a checkbox exercise.
Herron's framing aligns with a broader trend that analysts at Gartner have tracked: the rise of what they term "digital sovereignty" as a strategic priority for large enterprises. According to Gartner's research, by the mid-2020s a significant proportion of large organisations will have formalised sovereignty requirements into their cloud procurement criteria. For European IT services firms like Atos, which can demonstrate end-to-end data residency, jurisdiction clarity, and regulatory alignment across the EU's patchwork of national implementations of GDPR, this is a structural sales advantage.
"It's a good time to be a European company — the regulatory environment that once felt like a constraint is now becoming our calling card with enterprise clients who need certainty about where their data lives and who can access it."
— Michael Herron, CEO, Atos UK and IrelandThis dynamic is particularly relevant for small business owners and entrepreneurs building on top of cloud infrastructure. The emerging ecosystem of European cloud-native services — from storage providers operating under strict EU jurisdiction to open-source AI tooling developed under transparent governance — reflects a market responding to demand. Companies like Hetzner, OVHcloud, and Infomaniak have all grown significantly by positioning themselves as GDPR-native alternatives to AWS, Azure, and Google Cloud. Atos's enterprise-scale positioning occupies the same conceptual space, but targets larger institutional clients with complex hybrid and multi-cloud environments.
Navigating the Human-AI Agent World: What Atos's Approach Reveals
One of the most substantive themes in Herron's podcast discussion is Atos's strategic orientation toward what he describes as a "human/AI agent world" — an operational model where AI agents handle defined, repeatable tasks while human workers focus on judgement-intensive, contextual, and client-facing activities. This is not a novel concept, but hearing it articulated by the head of a major European IT services provider reveals something important about where enterprise AI deployment is actually heading, as distinct from where the press releases claim it is.
The human-AI collaboration model that Herron describes aligns closely with frameworks being developed by the European Commission's AI Office, which emphasises human oversight as a core requirement for high-risk AI applications. Under the EU AI Act, systems deployed in areas like HR decision-making, critical infrastructure management, and certain cybersecurity applications must maintain meaningful human control — a requirement that, by design, mandates the kind of hybrid human-agent architecture Atos is building toward.
For IT decision-makers evaluating AI tooling, this has practical implications. Agentic AI systems — where AI models can take autonomous actions, call external APIs, manage workflows, and interact with other AI agents — are moving rapidly from research prototypes to enterprise deployment. The challenge is governance: who is accountable when an AI agent makes a decision that has downstream legal or financial consequences? European regulatory frameworks are forcing this question into the open earlier than US-based enterprises tend to confront it. According to research published by McKinsey & Company on the future of work and AI, organisations that establish clear human-machine accountability frameworks earlier in their AI deployment cycle experience significantly fewer costly remediation events.
| AI Deployment Model | Human Oversight Level | EU AI Act Risk Category | Typical Enterprise Use Case |
|---|---|---|---|
| Fully Automated AI Agent | Minimal | High / Unacceptable | Autonomous hiring decisions, credit scoring |
| Human-in-the-Loop AI | Review at decision point | High (compliant) | Medical diagnostics assistance, fraud flagging |
| Human-on-the-Loop AI | Monitoring with override | Limited / Moderate | IT operations, infrastructure monitoring |
| AI Copilot / Augmentation | Human-led, AI-assisted | Minimal / Transparent | Code generation, document summarisation |
The Rising Cost of Powering AI: A Problem European Infrastructure Must Solve
Herron's discussion also touches on a concern that is increasingly front-of-mind for anyone running enterprise AI workloads: the surging energy costs associated with large-scale AI inference and training. This is not an abstract environmental concern — it is a direct operational cost driver that is reshaping how organisations evaluate cloud infrastructure choices and where they locate AI compute.
Training and running large language models is extraordinarily energy-intensive. According to reporting by Wired and corroborated by analysis from the International Energy Agency (IEA), data centre electricity consumption globally is projected to grow significantly as AI workload demand accelerates. For European enterprises, this creates a specific tension: the EU's carbon reduction commitments and energy cost structures mean that AI infrastructure decisions carry both financial and regulatory dimensions that their US counterparts may not face to the same degree.
For Atos, which operates significant data centre and managed services infrastructure across the UK and continental Europe, this creates both a challenge and an opportunity. European data centres that run on higher proportions of renewable energy — particularly in Scandinavia and Iceland, where hydroelectric and geothermal power is abundant — can offer AI compute that is both more carbon-efficient and, depending on energy market conditions, more cost-predictable than facilities in regions with more volatile energy markets. This is an area where European cloud infrastructure providers are actively competing on sustainability credentials alongside the traditional metrics of latency, uptime, and compliance.
Estimated renewable energy share for data
Originally reported by UKTN. Summarised and curated by European Purpose.
