Atos unveils Genesis Intelligence AI suite with Codex platform
Atos has announced the Atos Genesis Intelligence suite, a broad AI platform designed to take enterprises from data management through to production AI deployment within a single governed ecosystem. Launched in Paris on 26 October 2026, the suite targets organisations in highly regulated sectors including defence, critical infrastructure, public services and transportation that require strict data sovereignty and auditability.
At the centre of the announcement is Atos Codex, described by the company as the "culmination" of the Genesis Intelligence suite. Short for Centre of Operations and Data Excellence, Codex is positioned as an operational AI platform rather than a recommendation engine. The system builds what Atos calls an "operational ontology" - a structured representation of an organisation's entities, rules, relationships and permissible actions, through which AI-generated insights can be translated into governed, executable decisions with configurable levels of human oversight.
What the suite offers
Genesis Intelligence bundles several components: a data foundation layer covering ingestion, cataloguing, lineage and quality; a model development capability Atos calls SLM Factory, focused on small language models that can be deployed locally; secure AI orchestration with cost-optimisation routing; and an enterprise AI portal providing access to a marketplace of customisable GenAI agents. The suite is deployable across public cloud, sovereign cloud, on-premises and air-gapped environments, a configuration breadth designed to appeal to public-sector and defence buyers with strict data-residency obligations.
Philippe Salle, chairman and chief executive of Atos Group, said the suite enables customers to "accelerate AI adoption, reduce complexity and deliver measurable value faster across the full AI lifecycle." Atos did not disclose pricing, named reference customers, or benchmarks for Codex's decision-cycle performance at launch.
Market and competitive context
The sovereign AI market is intensifying as European governments and large regulated enterprises seek alternatives to hyperscaler platforms that route data through US or non-EU infrastructure. Competitors in the enterprise AI services and orchestration space include Capgemini, Sopra Steria, and to a degree IBM Consulting on the services side, alongside infrastructure-layer sovereign cloud providers such as OVHcloud and Deutsche Telekom's Open Telekom Cloud.
Atos Group, which generated revenues of approximately €7.2 billion and employs close to 53,000 people, has been rebuilding its commercial profile after a prolonged restructuring that saw it split its IT services and technology products businesses. The Genesis Intelligence launch is an attempt to reposition the Atos services brand around AI, drawing on Eviden's cybersecurity and mission-critical systems heritage. Whether the company can convert a well-specified product architecture into a pipeline of named enterprise wins will be the key test over the next twelve months.
Regulatory tailwinds
The suite's sovereignty and governance framing aligns closely with the direction of EU regulatory travel. The EU AI Act's high-risk AI provisions, which apply directly to AI systems used in critical infrastructure and public-sector decision-making, require transparency, human oversight and documented governance, all capabilities Atos claims Genesis Intelligence and Codex provide. Similarly, NIS2 obligations on critical-entity resilience reinforce the commercial case for on-premises and air-gapped deployment options that keep operational data within organisational control.
For defence and government customers, the company's existing security clearances and its operating model across 54 countries give it access to procurement routes that cloud-native AI vendors cannot easily match. Atos will need to demonstrate, through reference deployments and third-party validation, that Codex's operational ontology approach delivers measurable reductions in decision cycle times in live mission-critical settings.