Northern Ireland, Scottish and Welsh Legislative Consent sought. Relevant documents: 3rd Report from the Constitution Committee, 7th Report from the Delegated Powers Committee
Northern Ireland, Scottish and Welsh Legislative Consent sought. Relevant documents: 3rd Report from the Constitution Committee, 7th Report from the Delegated Powers Committee
12:15 pm
Clause 15: Reporting of incidents by regulated persons
17: Clause 15, page 21, line 19, leave out “capable of having” and insert “likely to have”
Member’s explanatory statement
This amendment would narrow the definition of “incident” in regulation 1(2) of the NIS Regulations so that it covers events likely to have an adverse effect, rather than events capable of having one, aligning the definition with the likelihood tests applied to notification later in the Bill.
17: Clause 15, page 21, line 19, leave out “capable of having” and insert “likely to have”
Member’s explanatory statement
This amendment would narrow the definition of “incident” in regulation 1(2) of the NIS Regulations so that it covers events likely to have an adverse effect, rather than events capable of having one, aligning the definition with the likelihood tests applied to notification later in the Bill.
My Lords, in moving Amendment 17, I will also speak to Amendment 28, which is closely related. Amendment 17 is in part a probing amendment about what constitutes an incident and the circumstances in which reporting is obligatory. It does not affect the amendment that I think the Government will move immediately afterwards.
As drafted, Clause 15 gives the very strong impression that an incident “capable of having” an adverse effect on security must be reported. If this is the case, it constitutes a much wider definition of what should be reported than if it were described as an incident “likely to have” an adverse effect. I think it is a widely held view—it is certainly the case in the industry and a point with which I agree—that “likely to have” would be far too wide a definition and would lead to extensive overreporting and an undue and unnecessary burden on regulators. Looking at the drafting, I asked myself what was the point of the “capable of having” definition in Clause 15.
I shall put forward a hypothesis. It would be very helpful if the Minister could confirm that it is a correct understanding of the existing draft, and that it does not mean that all incidents capable of having an adverse effect on security will need to be reported. Is it right to say that the definition in Clause 15 of what constitutes an “incident” applies across the whole of the regulations, and therefore feeds into security as well as reporting duties? That is to say, firms have a preventive duty to defend against what could be and what could happen, as well as what is likely to happen. That is a preventive duty. Can the Minister confirm that the phrase “capable of having” means that firms should have adequate preventive policies, but it is not—this is where the point comes in—the trigger for an incident to be reported, because in each case this requires it to have affected or be affecting the system?
My Lords, I will speak to this core group of amendments on incident reporting, in particular to Amendment 165, standing in my name, while addressing the other amendments in this group. First, Amendment 17, which was very cogently set out by the noble Baroness, Lady Neville-Jones, addresses what has emerged as one of the most contentious technical faultlines, in our view, across Part 2 of this Bill: the statutory threshold that triggers mandatory incident reporting to the designated competent authority, the NCSC. As the Bill is drafted, Clause 15 fundamentally widens the reporting net by redefining a reportable incident to include any event that is merely “capable of having” an adverse effect on the security of network and information systems, as the noble Baroness described.
While one can readily understand the cyber security community’s desire for complete visibility, in practice, the phrase “capable of having” is an operational disaster. In the daily reality of enterprise networking, thousands of automated port scans, routine phishing lures and perimeter firewall probes occur every hour. Almost every single one of these low-level events is technically capable of having an adverse effect, if multiple defensive layers were to fail simultaneously. By forcing businesses to notify regulators under threat of £17 million penalties whenever an event is merely “capable” of causing harm, the Government will unleash an administrative tsunami of defensive reporting.
Rather than enhancing national security, this compliance overload will drown NCSC analysts in background noise, making it far harder to detect sophisticated state-sponsored attacks. Amendment 17, in our view, would resolve this by replacing “capable of having” with the objective standard of “likely to have”. This would restore the established probability threshold used across UK regulatory frameworks, ensuring that mandatory notifications are reserved strictly for genuine material threats where there is a real likelihood of operational compromise.
12:30 pm
Amendment 28 in the name of the noble Baroness, Lady Neville-Jones—again, cogently introduced by her—removes “could have had” from the definition of a data centre incident, so that the definition instead covers incidents that have had, are having or are likely to have a significant impact. This closes off the same speculative hypothetical triggering that this Committee has already discussed under Amendment 17, earlier in the clause, and applies the same discipline specifically to data centres.
I turn to Amendment 171 from the Conservative Front Bench, on a ransomware reporting amnesty. I look forward to that being introduced. While we agree that incident reporting must be encouraged, an amnesty creates, in our view, a dangerous moral hazard, potentially letting negligent boards off the hook and indirectly legitimising ransom payments. We believe that the true solution is proactive safe harbours such as our presumption of conformity under Amendment 94, which we debated on Tuesday.
Regarding Amendment 172 and calls for a single reporting portal, industry does not need another 12-month review. The answer was already on the amendment paper under Amendment 88 from the noble Lord, Lord Birt, which we discussed on Tuesday, establishing an office for cyber resilience as the single national threat-reporting hub, which sadly did not receive the approval of the Conservative Front Bench or, indeed, the government Front Bench.
Regarding Amendment 172 and calls for a single reporting portal, industry does not need another 12-month review. The answer was already on the amendment paper under Amendment 88 from the noble Lord, Lord Birt, which we discussed on Tuesday, establishing an office for cyber resilience as the single national threat-reporting hub, which sadly did not receive the approval of the Conservative Front Bench or, indeed, the government Front Bench.
My Lords, I start by thanking my noble friend Lady Neville-Jones for introducing this group and setting out her stall so clearly and compellingly. I apologise that some of the amendments that have been looked at here I had in my record as being part of the next group. So, if I do not cover them all now, they will be covered by my noble friend Lord Markham as we get into the next group.
Let me begin by outlining the amendments in my name and those of my noble friends Lord Markham and Lord Holmes of Richmond. The need for action on ransomware has never been higher. The NCSC handled 204 nationally significant ransomware attacks in the year to September 2025 that we know about—up by 130% on the year prior, leading the NCSC to name ransomware as the most pressing threat to the country in its annual report. Of course, one of the challenges we face with ransomware attacks is not knowing when they happen, to whom and how often. The victims too often have strong reasons, generally associated with legal liability, not to report them. This makes it challenging, if not impossible, for any government agency seeking to identify commonalities across attacks to pursue repeat offenders and warn vulnerable organisations.
We could seek to make reporting of such attacks mandatory, but at the risk of placing hacked organisations in an impossible position where public reporting creates a legal bind that worsens the damage already done by the attack. I take on board the cogent concerns expressed by the noble Lord, Lord Clement-Jones, but the moral hazard occurs today where companies do not report ransomware attacks, thereby damaging our collective ability to defend others yet to be attacked.
Our amendment therefore seeks to find a channel that reports the facts of the hack and the metadata around it in a way that is not disclosed beyond the agency charged with cyber protection and does not become public knowledge. I do not pretend that this will be straightforward. For instance, we would have to understand how to deal with FoI requests and so on. That is why we propose a consultation. But if we were able to achieve something on this basis, we would greatly enhance our ability to protect UK PLCs from these hugely damaging attacks.
I thank noble Lords for their amendments in this group; in fact, subsequent groups also speak to this question of the nature, scope and timeliness of incident reporting. What we are all trying to do, I think, is to get the right balance in reporting actionable information that can be used by regulators and the NCSC to improve the security of the United Kingdom and the entities that operate essential services within it. That is obviously what the Government have put forward. I have heard clearly the arguments made by noble Lords, some of which probe the intention and the detail, and I will attempt to clarify those as I speak.
First, I shall speak to Amendments 19, 36 and 44 in my name. Improving incident reporting under the NIS framework is a key pillar of the Bill. Without an understanding of incidents, our regulators and the NCSC cannot assist in recovery, assess risk and bolster resilience. The amendments that I have tabled will ensure that the incident reporting measures for regulated entities reflect what we are trying to achieve.
The Bill already requires relevant regulated entities to consider a list of factors when determining whether an incident is likely to have a significant impact and be reportable. This includes whether data relating to users is, or is likely to be, compromised. Government Amendments 19, 36 and 44 remove the reference to “users”, meaning that all data compromises relating to the relevant network and information system are in scope of incident reporting. This will enable key incidents to be reported, including the compromise of commercially sensitive information or the exposure of access details or usernames of the regulated service.
These incidents will need to be reported to the NCSC and the relevant regulator. I say in response to the noble Lord, Lord Clement-Jones, that that is the motivation behind the change to that categorisation. This will ensure that the regulators have full oversight of significant security compromises, supporting them to keep the UK safe and secure. We will shortly consult on what constitutes a significant impact and put further detail in secondary legislation and guidance.
May I interrupt the Minister before she moves on to the next set of amendments? I do not intend to ambush her as regards her amendments this time around, but I seek an assurance, given that there seems to be quite a philosophical difference between her amendments today and those put forward by the noble Baroness, Lady Neville-Jones. There is considerable industry concern about the disproportionality involved. I seek an assurance from the Minister that, between Committee and Report, she will actively consult on the impact of this part of the Bill—Clause 15—and not just when it is in black-letter form. There is quite a lot of concern from many industry voices. It is incumbent on the Government to listen to those voices on the impact of this reporting structure and these duties before they go ahead in a way that many of us believe will not be helpful for the running of these businesses.
20 of 271 shown
I am making a distinction between “capable of having”, which applies to a duty to pursue preventive policies, and the trigger of the duty to report, which lies not in the phrase “capable of having” but in “likely to have”. Then there are examples of what I am saying in the regulations, and I can cite them: Regulation 11(3)(a), on page 21 at line 35; Regulation 12A(2)(a), on page 26; and Regulation 14E(2)(a), on page 29 at line 27. If the Minister can confirm that, within existing structures, what I have said is correct—there are no circumstances in which “capable of having” would be the reporting trigger—that would be a very helpful clarification. I will listen closely to the Minister’s reply on this point.
There is a “however”: there is a snag when it comes to the introduction of data centres, and that is the object of my Amendment 28. Data centres sit outside the existing structures that I have just talked about but, as yet in the drafting, there are no reporting trigger regulations for them. It is intended that the data centres should be, in future, big players in the system, so it matters that there is a gap in our information about the circumstances in which they would have a duty to report. It is an odd anomaly. New Regulation 11A(3)—on page 23, from lines 13 and 14 onwards—makes reportable
“an incident which could have had … a significant”
effect, whether or not it had any impact at all or anything was affected. As there is no list of factors for judging what constitutes a significant attack in the Bill, it makes it quite difficult to interpret.
For the operators of essential digital services and managed service providers, such factors are set out expressly in the new regulations in the Bill. However, they are absent for data centres. Why is this the case? What is the rationale for what appears an anomaly? It means that, when reporting an incident, a data centre has to do so when any of the following have had, or were likely to have,
“a significant impact on the operation or security of the network and information systems relied on to provide the data centre service … a significant impact on the continuity of the data centre service … or … any other impact, in the United Kingdom or any part of it, which is significant”.
These are very wide definitions of liability to report, and the discrepancy between them and those applying to other operators seems neither sensible from a security point of view nor fair for different business circumstances, as there will be all sorts of different businesses using data centres.
Although I hope that this will not be the case, I fear that the Government may say that the thresholds and factors for all categories of business will be set out in secondary legislation and subject to consultation. I ask the Minister to think hard about the adequacy of that reply. We are talking here about a penalty-backed duty, which is the core element of the Bill; it is not some minor point. It would seem a poor legislative approach in a foundational Bill for a new regime to fail to define the factors leading to a penalty for a significant segment of providers, when there are indicators in the Bill for other categories of provider. Those other players have different, less demanding and more sensible terms for a trigger for reporting. If data centre regulations need to be different from those for the other players that I have mentioned and the rest of the market, can the Minister explain why? It is the kind of complexity that will give the sectoral approach to regulation a controversial reputation, because it immediately raises the issue of making different rules for people who are apparently, in practice, in the same category. I hope that is not the case and that the issue can be resolved by remedying the drafting.
To sum up, in addition to my request for a clear statement from the Minister about the trigger for a duty to report in existing structures being related to the likelihood of an adverse effect on security and not on capability, I hope she will also take seriously the need to level the playing field for data centres on this issue and remedy what seems an important defect in the drafting of the Bill. I beg to move.
This issue is compounded by the Government’s own drafting amendments, specifically Amendments 19, 36 and 44, which replicate the ultra-broad definition of compromise throughout parts 2 and 3. By removing “users” from Clause 15 and redefining data compromise to cover any event affecting data stored or processed on a system, the Government are dramatically expanding the notification net to include technical data anomalies that cause zero destruction or loss to actual customers. Combining this sweeping definition of data compromise with the low “capable of having” trigger will hugely affect responsible operators. It will force critical suppliers and small digital providers to spend their limited resources filling in compliance paperwork, rather than actively defending their infrastructure.
We risk creating a reporting system that captures everything and understands nothing. We must have objective reporting thresholds. By accepting the noble Baroness’s Amendment 17, restoring the “likely to have” test, we would ensure that mandatory reporting delivers high-quality actionable threat intelligence, rather than an unmanageable flood of routine notifications.
Under the new reporting regime, hundreds of incidents will be notified to regulators and the NCSC, but at present the Bill lacks any mechanism to ensure that aggregate intelligence is shared with Parliament or industry. Under Amendment 165 in my name, I propose that the Government lay an annual anonymised report before Parliament, detailing incident volumes, sector breakdowns and principal attack vectors. This would provide software developers and CNI operators with the situational awareness needed to harden defences.
Amendment 172 seeks to require a review on the impact of the new reporting requirements introduced by the Bill. Again, this is fairly straightforward. The strengthened incident reporting requirements are being introduced to allow the regulators and the Government to help with providers and suppliers who have been attacked. Whether these requirements actually serve that purpose, and whether they do so at the expense of providers, cannot yet be known, but we must be able to form an assessment and adjust if necessary. Everyone in this Room would accept that we need statutory agility in the face of fast-moving technology, and a review on these lines could and would enable just that.
For a similar reason, I support the desire for transparency in Amendment 165 in the name of the noble Lord, Lord Clement-Jones. This may even overlap with our own amendment; we could probably think about merging the two in some way. It seems clear that both Houses of Parliament should be informed as to what the reporting regime is being used for and whether it is fulfilling its function. I hope that the Minister agrees.
I very much support Amendment 17 in the name of my noble friend Lady Neville-Jones. We are going from an incident constituting an actual adverse event on the security of network and information systems to it being capable of having such an effect. Arguably—the noble Lord, Lord Clement-Jones, made this point very well—almost any incident would meet this condition. We need language that expresses genuine risk to avoid all incidents being caught in the net. This seems wholly pragmatic to me and I commend it to the Minister, to whose response I look forward.