Sovereign Cloud Gap Analysis

Why Scaleway

Scaleway is a European cloud provider, owned by the French iliad group, with all data centres in the EU (Paris, Amsterdam and Warsaw). Because it falls only under European law, it is not subject to US legislation such as the CLOUD Act or FISA, which can give US authorities access to data held by US hyperscalers, even when that data is stored in Europe.

Scaleway is also closely tied to Mistral AI, Europe's leading AI model developer. Mistral trained models such as Mixtral on Nabu, Scaleway's AI supercomputer with more than 1,000 NVIDIA H100 GPUs. Mistral's models are available as managed endpoints through Scaleway's Generative APIs. Large European companies are already building on this combination like Airbus and ASML. Many of them chose Scaleway as the cloud for sensitive industrial and defence workloads.

Other European providers such as Hetzner and Clever Cloud were not considered, because they do not aim to be a full hyperscaler. Hetzner focuses on low-cost virtual machines and bare metal. Clever Cloud is an application platform (PaaS). Neither offers the broad set of managed services, such as Kubernetes, serverless, databases, IAM, key and secret management, and audit logging, that a platform comparable to GCP needs. Of the European providers, Scaleway comes closest to that ambition, which makes it the most meaningful comparison.

The question this analysis answers is whether that sovereignty can be had without giving up the security and governance controls we rely on in GCP today.

Summary

Scaleway has real strengths. Its serverless offering is mature: Serverless Containers, Jobs and Functions cover most workloads that run on Cloud Run and Cloud Functions today, with scale-to-zero and simple pricing. Networking is solid, with regional VPCs, private networks, load balancers, Public Gateways and site-to-site VPN. For AI, Generative APIs and Managed Inference offer managed endpoints for open models, hosted entirely in Europe, which is a strong sovereign option for AI workloads.

Its weaknesses lie in governance, security and compliance. IAM applications are always organisation-scoped, there are no custom roles or deny policies, and there are no resource-level policies or organisation policies to enforce guardrails. Workloads cannot get short-lived credentials, so static API keys are unavoidable. There is no managed equivalent of Security Command Center, Binary Authorization or Privileged Access Manager. Audit logging does not cover every product and has no data-access logs. Each of these gaps has to be filled with tooling we build and run ourselves, or accepted as risk.

That makes Scaleway a good fit for a single company with no need to separate departments and weaker compliance needs. Its flat hierarchy (organisation → project, no folders) and fixed permission sets work well when the blast radius stays within one team and no trust boundaries are needed. It is not yet a fit for large, regulated organisations that need strong isolation between departments, enforceable guardrails and audit-ready evidence. For them, building an enterprise-grade platform on Scaleway today would be an uphill effort.

Gaps

Topic GCP Scaleway Impact
Identity-Aware Proxy (IAP) In GCP, IAP provides zero-trust access. It authenticates and authorises requests before they reach backend services. No VPN needed — users access internal apps through Google's identity layer. No equivalent service exists. Scaleway has no managed zero-trust proxy offering. Without an native zero trust proxy solution the easy way to secure database access is an ip white list. There are self hosted iap solutions, but those come with some administrative burden.
IAM Application Scoping (Service Accounts) In GCP, Service accounts are project-scoped. You grant custom or predefined roles per-project, enabling strict least-privilege. Resource-level IAM policies are widely supported (GCS, Pub/Sub, KMS, BigQuery, etc.). Deny policies and org policy constraints provide guardrails. IAM applications (service account equivalent) are always organisation-scoped — they cannot be created in or confined to a project. Permission grants can be scoped to org or project level, but the application identity itself is org-level. No custom roles exist — only fixed Scaleway-defined permission sets. Policies are allow-only (no deny, no guardrails). Almost no resources have resource-attached IAM policies (Object Storage buckets are the only exception, and they are intersective with IAM, not additive). Resource-level CEL conditions exist for IAM, Key Manager, and Secret Manager only. Least-privilege is significantly harder. The application identity is always org-scoped, so policy is the sole boundary. Resource-level access control is largely unavailable outside Object Storage.
Workload Identity Federation In GCP, External workloads (GitHub Actions, GitLab CI, AWS/Azure, other Kubernetes clusters) authenticate via short-lived tokens (OIDC/SAML/X.509) through STS — no long-lived service account keys needed. No native equivalent exists in Scaleway. Scaleway's IAM API has no token-exchange, identity-pool, or STS endpoint. SAML/SCIM federation covers human users only. CI/CD pipelines must use long-lived access-key/secret-key pairs stored as CI secrets. Every external automation (GitHub Actions, GitLab CI, etc.) requires a static API key that we must mint, store, and rotate manually.
Workload Identity In GCP, Managed workload identities bind GKE pods, Cloud Run instances, and Compute Engine VMs directly to service accounts. The workload receives short-lived tokens via the instance metadata endpoint. Scaleway has no native vm/pod/container to IAM application binding. Pods must obtain credentials via Secret Manager or External Secrets, which relocate the key rather than eliminate it. The bootstrap key for the CSI/ESO controller itself is unavoidably long-lived — the secret-zero problem. Every workload ends up holding a static API key. Secret Manager CSI improves storage hygiene and enables central rotation a bit, but its still fundamentally long-lived and requires careful management.
Organization Policies In GCP, Organization Policy Service provides centrally-managed, hierarchy-inherited preventative constraints on how resources may be configured — independent of IAM. ~200 managed constraints (restrict resource locations, block VM external IPs, block Cloud SQL public IP, disable service-account key creation, domain-restricted sharing, GKE hardening) plus custom CEL constraints, with dry-run mode and Policy Simulator. Scaleway has no equivalent. Data residency, "no public IP / no public bucket", and credential-hygiene rules cannot be enforced server-side. They must be shifted left into the codebase using policy-as-code in CI. Enforcement depends on nobody holding direct create permissions outside the pipeline.
Binary Authorization In GCP, Binary Authorization only lets signed, trusted container images be deployed to GKE and Cloud Run. Unsigned images are blocked. It works with Cloud Build and Artifact Registry out of the box, and it keeps checking running workloads after they are deployed. Scaleway has no equivalent. Container Registry only stores images. It has no image signing or vulnerability scanning. We would have to build this ourselves, which leaves a gap for supply-chain compliance that GCP covers with a managed service.
Security Command Center (SCC) In GCP, Security Command Center gives one organisation-wide view of security risks. It finds misconfigurations (public buckets, open firewalls, overly broad IAM), flags vulnerabilities, detects threats such as crypto mining, leaked keys and suspicious API activity, and reports compliance against standards like CIS and ISO 27001. Scaleway has no equivalent. Audit Trail records who did what through the API, and Cockpit holds logs and metrics, but nothing analyses them or reports findings. There is no posture scanning, threat detection or compliance reporting. We would have no central view of our security posture. Misconfigurations and threats only get noticed if we build our own detection on top of Audit Trail and Cockpit logs, or bring in a third-party tool. Proving compliance to auditors becomes manual work.
Customer Identity (CIAM) In GCP, Identity Platform is a managed, serverless CIAM service (email/password, phone, social, anonymous, SAML/OIDC, MFA, multi-tenancy, blocking functions). Pay-per-MAU: the first 50,000 MAU on email/social/phone providers are free, then $0.0055/MAU and lower at higher volumes. SAML/OIDC: 50 MAU free, then $0.015/MAU. Scaleway has no native CIAM service. The option Scaleway points to is Cloud-IAM, a French managed Keycloak provider that runs on Scaleway infrastructure and is listed in the Scaleway Marketplace. Pricing goes up with the total number of stored user accounts, not MAU, and pricing for larger user counts is quote-only. CIAM goes from a pay-per-use service that is often free to a fixed monthly fee. At a realistic production tier (Essential or higher), Cloud-IAM costs at least about €6k–17k/year, where Identity Platform is €0 up to 50k MAU and about $275/month at 100k MAU. Because Cloud-IAM counts stored accounts, dormant users also cost money.
Audit Logging In GCP, Cloud Audit Logs cover every service. Admin Activity logs record every configuration change, and Data Access logs record who read or changed which data (objects, tables, secrets). Logs can be routed organisation-wide to a locked storage bucket or SIEM. Scaleway Audit Trail only logs management API calls, and not for every product. Object Storage, Block Storage, Container Registry, Billing and Domains/DNS are not integrated yet. There are no data-access logs. We cannot show who accessed which data, and changes to buckets, volumes and images leave no audit record. This weakens incident investigation and makes it hard to meet DORA and internal audit logging requirements.
Compliance Certifications In GCP, the platform holds ISO 27001/27017/27018, SOC 1/2/3, PCI DSS and many others. Audit reports can be downloaded on demand from Compliance Reports Manager. Scaleway lists ISO 27001, HDS (French health data), CSA STAR Level 1 (a self-assessment) and GDPR. It has no SOC 2 or PCI DSS, but this can be explained by the fact that these are American Standards. Some extra work is needed to map ISO 27001 and HDS to SOC 2 requirements and fill in the gaps.
Privileged & Provider Access In GCP, Privileged Access Manager gives just-in-time, approved and time-limited elevation to admin roles. Access Transparency logs when Google staff access our data, and Access Approval lets us approve or deny that access first. Scaleway has no equivalent. Admin permissions are standing grants, and there is no visibility into or control over access by Scaleway staff. Admins hold permanent high privileges, so a compromised admin account has full impact straight away. We would need our own break-glass and approval process. Access by the provider cannot be shown to auditors.