The Cabinet Office is creating a central insourcing unit to identify critical government services that could be brought back inside the public sector, adding institutional machinery to a wider attempt to reconsider when outsourced delivery offers the strongest value and operational control.

The unit will sit within the Office for the Prime Minister and Cabinet and coordinate work across departments, remove barriers to insourcing, and identify cross-government opportunities. It will also support efforts to reduce consultancy and professional-services spending by building capability inside the Civil Service.

The announcement ranges well beyond technology, covering areas including facilities management and other public-service contracts, but it reaches directly into digital government because large outsourced arrangements frequently combine technology, operations, specialist staff, data, and institutional knowledge rather than providing a narrowly defined piece of software.

Its creation follows the Public Interest Test introduced for most central-government service contracts valued above £1 million. Departments are expected to consider whether services should be delivered internally, bought from suppliers, or operated through a mixed model rather than treating one form of delivery as the automatic default.

Insourcing is a capability question

Changing the commercial model does not by itself create the skills required to deliver a service. Government can terminate or allow a contract to expire relatively quickly, but rebuilding architecture knowledge, software engineering, product management, cyber security, commercial expertise, service operations, and specialist domain knowledge can take much longer where those capabilities have sat with suppliers for years.

That creates a particular challenge in technology because outsourced services often contain a mixture of people, proprietary tooling, software licences, subcontractors, cloud services, documentation, and tacit operational knowledge. Transferring responsibility without understanding those dependencies can replace one commercial problem with an operational one.

The government’s sourcing approach recognises that difficulty by asking departments to consider value, control, resilience, and whether internal delivery is viable. In practice, many services are likely to remain hybrid because specialist suppliers can provide expertise that would be inefficient for government to maintain permanently, particularly during migrations, unusual security work, or temporary capacity spikes.

Recent Whitehall technology programmes show how blurred the boundary has become. A government technology and AI training framework worth up to £456 million was structured partly to support stronger internal Civil Service capability, leaving external suppliers involved in building some of the skills intended to reduce long-term dependency.

That is not necessarily contradictory. External specialists can provide temporary capacity while permanent internal teams retain ownership of architecture, data, product decisions, commercial strategy, and long-term service knowledge. The failure point comes when the customer loses enough understanding of the service that changing supplier becomes operationally dangerous.

Supplier dependence is not the same as outsourcing

The sharper question for digital government is therefore not simply whether external companies are involved but whether departments can change supplier, alter a system, or take a service back without losing control of the operation. A contract can appear competitive while leaving the customer heavily dependent on proprietary tooling, undocumented integrations, or expertise concentrated in one supplier team.

Conversely, a nominally in-house service can remain deeply dependent on external cloud platforms, enterprise software, contractors, and managed infrastructure. Insourcing frontline staff or service management may improve accountability while leaving most of the technical supply chain untouched.

The new unit will consequently need to distinguish between visible headcount and underlying capability. Bringing a function onto the public payroll has limited strategic value if architectural decisions, critical data knowledge, or the ability to maintain the system remain outside the organisation.

Cost comparisons can be equally difficult. Outsourcing can convert permanent staffing costs into contracts and provide access to specialist resources, but procurement, contract management, supplier margins, transition work, and change requests carry their own costs. Insourcing can remove some of those layers while creating recruitment, training, management, and technology expenses that have to be assessed over several years rather than only when a contract ends.

A cross-government unit could nevertheless expose repeated patterns that are harder to see when departments act individually. Common forms of supplier dependence, duplicated contracts, or shortages of specialist skills become easier to address when government can examine them across organisations rather than one procurement at a time.

The approach could also improve the state’s negotiating position where services remain outsourced. A department with credible internal expertise and a realistic option to take work back is less dependent on an incumbent than one whose operational knowledge disappears when the supplier leaves.

The first tests will come as existing contracts approach renewal. For digital services, those decisions will require more than a procurement-policy change: government will need enough technical knowledge to understand the systems it owns, enough commercial capability to manage the suppliers it still needs, and enough institutional memory that changing a contract does not mean relearning how a public service works.