(4 days ago)
Public Bill Committees
Kanishka Narayan
Clause 4 of the Bill amends the NIS regulations by creating a new regulated sector, data infrastructure, and designating the Secretary of State for Science, Innovation and Technology and Ofcom as joint regulators. We have received clear feedback from the data infrastructure sector expressing concerns that a dual regulator model could create unnecessary complexity and limit accountability. Amendments 11 and 12 will remove the Secretary of State for Science, Innovation and Technology as a regulator, leaving Ofcom as the sole regulator, which will streamline the regulatory model for data infrastructure and resolve the concerns raised by stakeholders.
Ofcom already has proven regulatory expertise and is well placed to oversee the new data infrastructure sector effectively. By adopting a single regulator for data infrastructure, the amendments will reduce administrative burden, simplify engagement, and strengthen accountability. This will ensure a clearer, more effective regulatory framework for this rapidly growing sector.
Clause 4 brings qualifying data centre services into the scope of the NIS regulations, recognising both their vital role in underpinning our economy and public services, and that disruption to them can significantly impact productivity, service delivery, and revenue.
Alison Griffiths
Clause 4 relies heavily on capacity as the trigger for regulation. I understand why that is attractive: it is measurable. But capacity is not the same as criticality, and a high-capacity facility used for redundancy can present less systemic risk than a smaller, highly concentrated one. I simply put on record that the way this threshold is applied in practice will matter more than the number itself.
Kanishka Narayan
I thank the hon. Member for that thoughtful point. One assurance I will offer her is that the direct definition of data centres in scope here rely on capacity as a proxy for their essential independent nature, but when data centres below the capacity threshold but high on the criticality threshold are suppliers to essential services, they would be covered in part by the critical suppliers framework in the Bill. I take her point into account.
Alison Griffiths
Clause 7 is definition-heavy, and rightly so; these terms decide who is regulated and who is not. My only observation is that cloud models are, as the Minister knows, evolving quickly because of the AI revolution. Definitions that track architecture too closely will age fast, so the Committee should be alert to whether these terms will still make sense in five years’ time and not just today.
Kanishka Narayan
I very much welcome that point. In talking about broad architecture characteristics—being able to scale compute and to be elastic to multi-tenants by being shareable—rather than setting out the specific nature of resources, we capture both commercial cloud and AI deployments. However, I am keen to ensure that we keep this under review and, where possible, use the flexibilities provided by the Bill to adapt it to changes in technology.
Although the policy intention behind the definition has not changed, amendment 13 will provide certainty for industry, support effective regulatory oversight and ensure that services whose disruption could significantly impact the UK economy and society are properly captured. In addition, the drafting is more aligned with that of our international partners, which will improve efficiency for providers operating across borders.
This targeted, technical improvement will bring greater clarity, consistency and fairness to the NIS regulations. I urge Members to support both the clause and this important amendment.
(4 days ago)
Public Bill Committees
Kanishka Narayan
What a pleasure it is to serve with you in the Chair. Clause 9 brings large and medium-sized managed service providers—MSPs—into the scope of the Network and Information Systems Regulations 2018. MSPs are organisations that provide an ongoing IT function, such as an IT help desk or cyber-security support, to an outside client. In doing so, MSPs often have widespread and trusted access to clients’ networks and systems. A single targeted attack can ripple outward, disrupting thousands of other systems. That makes MSPs attractive targets for cyber-attacks. Last year an attack on Collins Aerospace halted check-in and boarding systems at major European airports, causing international disruption. Such attacks highlight what can happen if a single point of failure is compromised, and the importance of managed service providers implementing robust cyber-protections. Despite that, MSPs are not currently regulated for their cyber-security in the UK. As organisations rely more and more on outsourced technology, we must close that gap. The clause provides essential definitions of a “managed service” and of a “relevant managed service provider” to clearly set out which organisations are in scope of the regulations.
Clause 10 imposes new duties on MSPs that have been brought into scope by clause 9. For the first time, such businesses must identify and manage risks posed to the network and information systems that they rely on to provide their managed services. As part of that duty, MSPs must have
“regard to the start of the art”,
meaning that they must consider new tools, technologies, techniques and methods that threat actors may employ. That includes artificial intelligence, and means that providers must deploy the right tools to mitigate the risks and take action to minimise the impact of incidents if they occur. By bringing MSPs into scope of the regulations and imposing such security duties on them, we will strengthen cyber-security and resilience across supply chains, reduce vulnerabilities in outsourced IT services and better protect businesses and services across the UK.
Alison Griffiths (Bognor Regis and Littlehampton) (Con)
Bringing MSPs into scope is the right direction of travel, and MSPs sit at points of concentrated risk, but they are not all the same and the real risk is not size alone but the level of privileged access and cross-customer dependency. Proportionality will be critical under these provisions if we want better security, not just box-ticking.
Kanishka Narayan
I agree very much with the hon. Member’s point, and a similar sentiment is expressed elsewhere in the Bill, in that it ensures that the focus is primarily on large and medium-sized MSPs, and that small businesses and microbusinesses are dealt with in a deeply proportionate way. That is an important point to take into account.
Clause 11 defines what it means for a digital or managed service provider to be
“subject to public authority oversight”
under the NIS regulations. Public authority oversight is defined as “management or control” by “UK public authorities” or by a board where the majority of members are appointed by those authorities. Such MSPs are already subject to requirements in the Government cyber-security strategy, which is mandatory for Government organisations. That ensures that cyber-resilience standards remain strong for services linked to public functions, while preventing disproportionate burdens on providers already subject to public authority governance.
In response to points raised by hon. Members in prior Committee sittings, I flag the engagement that we have conducted in coming to the definition of MSPs in question. In particular, beyond the provisions of the 2022 consultation, prior to the introduction of the Bill, we conducted a range of bilateral meetings. We have had multiple conversations with the industry body techUK, roundtables with digital firms, and we engaged through the National Cyber Security Centre-led MSP information exchange with 40 providers in this context, and undertook market research mapping the MSP market. As a consequence, adjustments to the definitions at the heart of this provision have been agreed with incredibly deep and broad engagement across the industry to arrive at a widely-welcomed definition.
Kanishka Narayan
Clause 12 will introduce a new power for regulators to designate critical suppliers to organisations as in scope of the NIS regulations. These are suppliers that are so pivotal to the provision of essential digital or managed services that a compromise or outage in their systems can cause a disruption that would have serious cascading impacts for our society and economy; I am thinking in particular of the Synnovis incident in 2024, when 11,000 medical appointments were cancelled across London hospitals as a result of an attack on a pathology service provider.
The clause will ensure that the power to designate can be exercised only where suppliers pose a credible risk of systemic disruption and when the regulator has considered whether the risks to the supplier cannot be managed via other means. In other words, it is a very high bar indeed.
The clause provides safeguards for suppliers, which must be consulted and notified during the designation process. It also requires regulators to consult other relevant NIS regulators when they are considering whether to designate, or decide to do so, ensuring that they have an accurate understanding of how suppliers are already regulated.
Finally, the clause provides for designations to be revoked when risks no longer apply or when a supplier has met the thresholds for regulation as a relevant digital service provider or relevant managed service provider. It should be noted that the clause does not set out the security duties on critical suppliers; these will be defined in secondary legislation following an appropriate period of consultation.
By addressing supply chain vulnerabilities, this measure will strengthen the resilience of the UK’s essential and digital services on which the public rely every day. I commend the clause to the Committee.
Alison Griffiths
The clause merits close scrutiny, because it is the point in the Bill where risk is supposed to be addressed beyond the individual operator and into the supply chain. In plain terms, clause 12 will allow the regulator to designate a supplier as critical where disruption to that supplier would have a significant impact on the delivery of an essential or digital service. The trigger is impact, not size or sector. That approach is sensible, but I want to stress-test how it works in the context of operational technology.
Across power, telecoms, transport, water and industry, many essential services rely on the same family of industrial control equipment. Substations, signalling systems and industrial plants may look different, but they often run on identical controlled devices and firmware supplied by a very small number of manufacturers.
The risk is not hypothetical. A single vulnerability in widely deployed OT equipment can create a common mode failure across multiple sectors at the same time, even where each operator is individually compliant with its duties. At the moment, the Bill places obligations squarely on operators of essential services, but in OT environments, operators do not control the design of equipment, the firmware, the vulnerability disclosure process or the remote access arrangements that vendors often require as a condition of support.
As Rik Ferguson highlighted in written evidence to this Committee, uncertainty about how and when suppliers might be brought into scope can lead to defensive behaviour and late engagement. The risk is amplified in OT, where suppliers may discover vulnerabilities before operators do, and where one operator may report an issue, while others in different sectors, using identical equipment, remain unaware.
There is also a traceability problem. OT equipment is frequently sold through integrators and distributors. Manufacturers may not have a clear picture of where the equipment is ultimately deployed. Without that visibility, national-scale vulnerability notification and co-ordinated response become very difficult.
UK Finance has also drawn attention to the complexity of multi-tier supply chains and the need for clear accountability when regulatory reach extends upstream. The clause recognises that reality, but its effectiveness will depend on how consistently and predictably designation decisions are made across sectors.
My concern is not about the existence of the power. It is about whether, in practice, the power will be used early enough and clearly enough to address shared OT risks before they become cross-sector incidents. Operational resilience today depends less on individual sites and more on the security practices of a relatively small— I would say very small—number of OT suppliers that sit behind them. The clause has the potential to address that, but only if its application is focused on genuine systemic risk and supported by clear signals to suppliers and operators alike. For those reasons, the clause warrants careful consideration as the Bill progresses.