Of all the obligations Chapter VI of the EU Data Act imposes on cloud providers, the rules around switching fees are the most commercially concrete and the most immediately actionable. Unlike the functional equivalence and data export requirements — which require engineering work and documentation updates — the switching fee rules require changes to pricing models, contract templates, and commercial terms that are already in market. The timeline is specific and non-negotiable: from September 2025, switching fees must be reduced to cost level; from January 12, 2027, they must be eliminated entirely.

For US cloud providers with European customer bases, this creates a defined compliance roadmap with real financial implications. Switching fees are, in many business models, a meaningful revenue line or a deterrent that supports customer retention. The Act eliminates them not because the European Commission doubts that providers incur costs during switching processes, but because it concluded that switching fees have been used as a market power tool — a structural barrier to competition that locks customers in regardless of their preferences. Understanding exactly what the Act requires, what it prohibits, and how to bring existing commercial arrangements into compliance is essential for any provider operating in the EU cloud market.

What Counts as a Switching Fee

The Data Act’s definition of switching fees is broad. A switching fee is any charge, financial or otherwise, that a cloud provider imposes on a customer in connection with the customer’s decision to move to a competing provider, move to an on-premises deployment, or otherwise terminate or reduce their use of the provider’s services in favor of an alternative. This definition captures the full range of financial mechanisms that providers have historically used to make switching expensive.

Data Egress Fees

Data egress fees are charges for transmitting data out of a cloud provider’s infrastructure. In cloud markets, data egress has historically been priced very differently from data ingress: it typically costs much more to move data out of a cloud than to move it in. While egress fees are charged for all outbound data transfers — not just switching-related transfers — any portion of an egress fee that is triggered by or associated with a customer’s decision to switch providers falls within the switching fee prohibition. Under the Act, the cost of egressing data as part of a switching exercise must eventually reach zero; in the intermediate period from September 2025 to January 2027, it must be reduced to cost level.

Service Termination Charges

Early termination fees, cancellation fees, or similar charges imposed when a customer ends a cloud service contract before its term expires are switching fees to the extent they are associated with the customer moving to a competing provider. A contractual provision that requires a customer to pay a penalty for ending a subscription early — when the reason for ending it is to switch to a competitor — falls within the prohibition. Providers cannot structure termination charges as something other than ‘switching fees’ to argue that the prohibition does not apply; the Act looks at the economic function of the charge, not its label.

Configuration Porting Charges

Any fee charged for exporting, porting, or documenting the configurations, schemas, automation scripts, security policies, or other technical settings that a customer needs to recreate their environment at a new provider is a switching fee under the Act. This category is important because it captures fees that might not superficially appear to be ‘switching fees’ but that serve the same economic function: making switching more expensive.

Other Charges Associated with Moving to a Competing Provider

The Act uses broad language to cover any charges ‘associated with’ moving to a competing provider. This is designed to prevent providers from repackaging switching fees as something else — for example, as an ‘account administration fee’ or a ‘data retrieval service fee’ — in order to avoid the prohibition. If the economic reality is that the charge is imposed because the customer is leaving, it is a switching fee regardless of how it is labeled.

The Two-Phase Timeline

The Data Act phases out switching fees in two stages. Understanding what each stage requires — and what it means to be ‘compliant’ at each stage — is essential for providers building their implementation plans.

Phase One: September 2025 to January 12, 2027

From September 12, 2025 (when the Data Act became applicable), switching fees must be reduced to cost level. This means that a provider may still charge for switching-related activities, but it may only charge the actual, documented cost of performing those activities. It cannot add a margin, it cannot charge a premium for the inconvenience of the customer leaving, and it cannot maintain pre-existing fee levels that exceeded cost recovery.

Determining ‘cost level’ for switching activities requires a genuine cost analysis. Providers must be able to document what it actually costs them to execute a switching request: the staff time involved in processing the request, the infrastructure costs associated with data export, any third-party costs incurred in connection with the transfer, and similar direct costs. A switching fee that is set at a level that cannot be justified by documented costs is not a cost-level fee and does not comply with Phase One.

This means that for providers currently operating under the Phase One requirement, simply reducing existing switching fees by a percentage is not an adequate compliance response. The correct approach is to conduct a cost-analysis exercise, document the actual cost of switching activities, and set fee levels accordingly. For most providers, this exercise will result in fee levels that are substantially lower than current market rates for egress and switching-related services.

Phase Two: January 12, 2027 Onward

From January 12, 2027, switching fees must be eliminated entirely. This is a hard prohibition, not a cost-level cap. After this date, a cloud provider subject to the Act may not impose any charge on a customer in connection with the customer’s decision to switch to a competing provider, switch to on-premises infrastructure, or otherwise move their data, workloads, and configurations away from the provider’s platform.

The January 2027 deadline is now less than two years away from the date of this writing, which means providers should already be planning for this transition. Changing pricing structures, updating contract templates, communicating changes to existing customers, and revising financial models that depend on switching fee revenue all take time. Providers that begin this work now are in a better position than those that wait until the deadline is imminent.

What Is Permitted vs. Prohibited

Drawing a clear line between what is permitted and what is prohibited under the switching fee rules is important for compliance planning.

Permitted, through January 2027: Charging a fee for switching activities, provided the fee is based on documented cost and does not include any margin or premium. Permitted indefinitely: Charging for cloud services that the customer continues to consume — the prohibition is on switching fees specifically, not on the ordinary pricing of cloud services. Permitted: Charging for data egress that is unrelated to switching — for example, egress charges for data transfers that are part of the customer’s ongoing operations, not part of a migration to a competing provider.

Prohibited from September 2025: Switching fees set at levels that exceed documented cost. Prohibited from January 2027: Any switching fee, regardless of its level. Prohibited regardless of timing: Structuring fees in ways designed to circumvent the prohibition by relabeling switching fees as something else, or by imposing fees on activities that are functionally equivalent to switching even if they are not labeled as switching-related.

One area of potential ambiguity is the treatment of minimum commitment contracts. Many cloud providers sell services under annual or multi-year commitments with pricing discounts tied to the commitment. If a customer exits a commitment early to switch providers, is the provider entitled to collect the remaining committed spend? This question is legally nuanced. A minimum commitment is not, on its face, a switching fee — it is a pricing structure that the customer agreed to. However, if minimum commitments are structured in a way that functions as a financial penalty for switching, they could attract regulatory scrutiny. Providers should review their commitment structures with EU counsel to assess the risk.

Auditing Existing Customer Contracts

For US cloud providers with established EU customer bases, one of the most urgent practical steps is auditing existing contracts for switching fee provisions that are now non-compliant or will become non-compliant by January 2027.

A contract audit for this purpose should examine several elements. First, identify any provisions that impose fees for data egress in connection with migration or termination. Second, identify any provisions that impose early termination fees or cancellation penalties. Third, identify any provisions that require payment for data export, configuration documentation, or API access in connection with ending the service relationship. Fourth, identify any provisions that create technical or administrative delays in the switching process that could be interpreted as a contractual mechanism for imposing switching costs.

Once non-compliant provisions are identified, providers face a question about how to address them. For new contracts, the solution is straightforward: update the standard template to remove or revise the non-compliant provisions. For existing contracts, the approach depends on the contract’s amendment provisions, the customer relationship, and whether the non-compliant provision is material enough to warrant a proactive outreach versus being addressed at the next renewal.

Providers should document their audit process and their remediation decisions. If a supervisory authority later investigates a complaint about switching fees, having a clear record that the provider identified and addressed non-compliant provisions proactively is a meaningful mitigating factor.

What Needs to Change in Standard Cloud Agreements

Standard cloud service agreements for EU customers — or agreements that include EU law choice of law provisions and govern services delivered to EU-based customers — need to be reviewed and updated to comply with the switching fee requirements. The specific changes will vary by provider, but there are common elements that most major cloud agreements will need to address.

Egress Fee Provisions

Provisions that impose data egress fees need to be reviewed to distinguish between ongoing operational egress (which may continue to be charged) and switching-related egress (which must be reduced to cost by September 2025 and eliminated by January 2027). If a provider’s egress pricing is uniform — the same fee applies regardless of whether the egress is operational or migration-related — the contract language may need to be updated to create this distinction, and the pricing schedules may need to be adjusted accordingly.

Termination Provisions

Early termination and cancellation provisions that impose financial penalties must be reviewed to ensure that any penalty associated with the customer’s decision to switch is removed or restructured. Providers may retain provisions that protect against early termination in contexts unrelated to switching — for example, customer-caused terminations for reasons other than moving to a competitor — but the legal boundary between these categories needs to be clearly defined in the contract language.

Portability and Export Provisions

The contract should include affirmative provisions confirming the customer’s right to export their data and configurations in machine-readable formats, through documented APIs, and without the imposition of fees beyond what the Act permits. This is both a substantive compliance requirement and a commercial signal to customers that the provider takes the Act’s portability obligations seriously.

Process Provisions

The contract should describe the switching process: how a customer initiates a switching request, what information they need to provide, what timelines apply to the provider’s response, and how the post-switch data retention period works. Having these provisions in the contract creates clarity for both parties and helps demonstrate compliance with the cooperation obligations in Chapter VI.

How Enforcement Is Likely to Proceed

The Data Act designates Member State competent authorities as the primary enforcement bodies for Chapter VI. Each Member State must designate authorities with the power to investigate complaints, conduct market surveillance, and impose administrative penalties. The Act sets maximum fine levels but allows Member States to calibrate enforcement within those limits.

Enforcement of the switching fee rules is likely to proceed in two ways. First, supervisory authorities may conduct market surveys to assess whether major cloud providers operating in their jurisdiction are complying with the cost-level requirement and, after 2027, the elimination requirement. These surveys may look at publicly available pricing information, sample contracts, and other documentary evidence. Second, individual customers who believe they have been charged non-compliant switching fees may file complaints with national authorities, which would trigger an investigation into the specific provider’s practices.

Early enforcement is likely to focus on the most visible cases: large providers with well-documented egress fee structures, and cases where a customer can demonstrate that they were charged a switching fee above cost level after September 2025 or any switching fee after January 2027. US hyperscalers with large EU market shares are particularly visible targets, and their public pricing pages and standard contract templates will be easily scrutinized by enforcement authorities.

The risk of enforcement is compounded by the fact that the Data Act’s switching fee rules are directly relevant to the concerns that prompted the European Commission to regulate cloud markets in the first place. This is not a peripheral provision of a complex regulation — it is one of the Act’s flagship obligations. Supervisory authorities will face political and policy pressure to enforce it visibly.

The Risk of Leaving Switching Fees in Contracts After 2027

A cloud provider that continues to impose switching fees after January 12, 2027 faces legal exposure on multiple fronts. Administrative fines from national supervisory authorities are the most obvious risk, but they are not the only one. Customers who are charged non-compliant switching fees may seek repayment of those fees under national contract law or through civil litigation. If a supervisory authority finds that a provider has systematically imposed non-compliant switching fees across its customer base, the financial exposure could be substantial.

There is also a reputational risk. The switching fee prohibition is well understood in the EU cloud market. Enterprise customers with legal and procurement teams will know what the Act requires. A provider that continues to impose fees that are clearly prohibited under a law that has been in effect for years will face commercial pressure as well as regulatory pressure: customers will cite the violation as a reason to seek alternatives, and the violation will be a negotiating point in contract discussions.

Providers should also be aware that the prohibition applies to all forms of the restricted charges, not just the most obvious ones. A provider that eliminates its labeled ‘egress fee’ but restructures its pricing to achieve the same economic effect through a different mechanism — for example, by introducing a ‘data retrieval service’ charge that applies only when a customer is in the process of migrating — will not be in compliance. The Act looks at economic substance, and enforcement authorities are unlikely to accept formalistic distinctions that preserve the commercial reality of a switching fee under a different name.

Planning for the Transition

The transition from the current Phase One regime to the Phase Two elimination of all switching fees requires deliberate planning across multiple functions. Pricing and finance teams need to model the revenue impact of eliminating switching-related fees and determine whether any offsetting adjustments to the base pricing structure are appropriate. Legal and contracts teams need to update standard agreements and manage the amendment process for existing contracts. Product and engineering teams need to ensure that the capabilities required to facilitate zero-cost switching are fully operational before January 2027. Customer-facing teams need to communicate the changes proactively, particularly to enterprise customers who are in long-term contracts that may currently include non-compliant fee provisions.

Providers that approach the January 2027 transition as a compliance exercise will minimize disruption. Providers that treat it as an afterthought will face a compressed remediation timeline and heightened regulatory risk precisely at the point when enforcement attention on the switching fee prohibition is likely to be highest.