Automated and Electric Vehicles Bill (Fourth sitting) Debate
Full Debate: Read Full DebateOliver Letwin
Main Page: Oliver Letwin (Independent - West Dorset)Department Debates - View all Oliver Letwin's debates with the Department for Transport
(7 years, 1 month ago)
Public Bill CommitteesI remind the Committee that with this we are discussing amendment 10 in clause 2, page 2, line 6, at end insert—
“or by an automated vehicle when transitioning between driving itself and being driven by a person,”
This amendment would ensure that the liability for accidents caused by an automated vehicle that is transitioning between driving itself and being driven by a person would be the same as the liability for accidents caused by an automated vehicle when driving itself.
My right hon. Friend the Minister rightly admonished me earlier in our proceedings for not making clear right from the beginning how the remarks I was making related to the structure of the Bill as it is and how it is trying to make progress without trying to solve all the problems.
In responding to the amendment of the hon. Member for Eltham, I want to ensure that I make clear why I am raising the point that I am raising about the Bill as drafted. I take it that the point of clause 2, which is one of the major points of the Bill, is precisely to ensure that the insurance industry has a clear and legally certain basis for proceeding. That is a restricted but very important ambition. The point that the hon. Gentleman raises in his amendment is very material from the point of view of realising the Minister’s ambition.
The way that the Bill is constructed, without the hon. Gentleman’s amendment or something like it, does not provide certainty for the insurance industry. The insurance industry has failed to recognise that the Bill does not provide that certainty. When the industry realises that it does not, it will blame us and the Minister for that and say, “Why on earth did you not give us certainty?” My whole intent is to ensure that the Minister can do what he is trying to do. I hope he will accept what I am saying in that light.
We had an interesting exchange in the course of the moving of the amendment about tier 3 and tier 4. To tell the truth, I do not have any faith in the tiers. They are a figment of a group of manufacturers’ imaginations. They are as good as we are going to get at the moment as a broad description of how things will go, but it is likely that all sorts of different things will be produced that are variously describable as tier 3-plus and tier 4-minus and God knows what else. I think the Minister has already agreed with what I think is certainly a true proposition: there will be at least a period in which people are experimenting with kinds of automation that involve significant opportunities for transition between the machine and the person. For that purpose, it does not matter whether we are talking tiers 3, 3-plus, 4-minus, 4 or, indeed, 4-plus.
There will possibly come a moment when drivers just fall out of the equation and there are not any drivers any more, just machines that take us to where we programme them to go. At that halcyon moment, probably decades from now, clause 2 would work fine, but the problem is that it will not work fine during what is likely to be the very long passage where there is a rather messy scene of vehicles that in varying circumstances are taken over by a driver or handed by the driver to the automation system. We were told in the evidence sessions with great certainty that it would take 10 seconds or less to hand over. We were also told that if a failure in the handover from the machine to the person occurred, all was well because the machine would find a way of stopping itself. I have learned, as I expect many members of the Committee have, always to take with a strong pinch of salt any assertion by assertive technologists that they know exactly how long it will take for something technological to happen in all circumstances. They do not know any such thing; they are speculating. They may prove to be entirely right—they certainly know a lot more about it than me—but it is perfectly possible that they will prove to be completely wrong.
The hon. Member for Eltham raised one circumstance in which the technologists could be very wrong. It may well be that the machines are so designed that they go to great lengths to wake up drivers who have gone to sleep when they have stopped driving and handed over to the machine. There may be rules enforced that say they must not go to sleep, but human beings are human beings, and they might go to sleep and it might take a lot longer than 10 seconds to wake them up. I happen to be married to someone who takes a lot longer than 10 seconds to wake up; I have no reason to suppose that every human being sitting next to the machine is going to be in full functioning order in 10 seconds. There could be quite long periods during which that transition is occurring.
The reason I say all that to my right hon. Friend the Minister is that we are not here talking about angels on pins; we are not talking about milliseconds that are just a figment of legal imagination. It is quite likely that, in real life, there will actually be some accidents that occur during periods of transition between machine and mankind. There is no reason we should be afraid of that; there are plenty of accidents on our roads now, and we are not entering into a new terrain in which there will be thousands more accidents—probably there will be thousands fewer. Nevertheless, some accidents might occur during transition. The Bill currently contains a binary choice. Either, as in clause 2(1),
“an accident is caused by an automated vehicle when driving itself”
or it is not. There is no allowance for the possibility of transition.
If a piece of legislation does not admit of a possibility, and that possibility comes about in real life and there is a court action about it, the court looks at the statute and it says to itself, “Blow me down! Once again, Parliament has been extremely stupid. There is nothing in the statute about this situation.” What does an English court do, thank goodness, under such circumstances? It invents the law. That is what it will do. It is not the case that there is a sort of legal black hole. Where there is statute and statutory construction does not lead to the answer to the case, the judge will invent the answer.
I take it that my right hon. Friend is speaking about fault. In those circumstances, what would be at question is where fault lies and what caused the accident. If that is the case, I direct him, without wishing to engage in a long debate about it, to clause 3(1), which deals with partial responsibility and therefore fault.
No, I am not raising the question of fault. I am raising the question of legal certainty about the circumstance. Clause 2 says that if the
“accident is caused by an automated vehicle when driving itself”
it is clear that
“the insurer is liable for that damage.”
It is equally clear, therefore, as a binary choice, that if the vehicle is not being driven by the vehicle itself, but by the driver, the driver is liable. Those two positions are perfectly clear. The insurer of the driver, who may or may not be a separate body from the insurer of the vehicle, takes on responsibility when the driver is driving. We are dealing here with the situation in which some combination of driver and vehicle has been the cause of the accident, during a transitional period from one to the other. The question arises, which of the two insurance policies is the relevant one? I do not believe that there is anything in clause 3 that solves that problem. If the Minister can point out something about the wording of clause 3, I hope you will allow him to do so, Mr Bailey, because it is definitely relevant to the point that the hon. Member for Eltham and I are raising.
My own view is that there is nothing in clause 3 that solves the problem, and therefore the courts will invent a solution. There is nothing wrong with that in general—the courts are very wise and may come up with a perfectly good solution—but the Minister’s purpose is not to say, “Let the courts invent a solution”. If that was his purpose, he would not need the Bill in the first place, because we have a common-law system. If there were no Bill, and if automated vehicles were to proceed and things were to go to court, the courts would find a solution. We would not need the Bill in the first place, if we were going to rely on the courts. The reason for having the Bill is to create legal certainty so that we are not simply trying to find out later, ex post, what the courts will make the law be. We are trying to make the law in advance, so that the insurance industry and the automated vehicle industry know how it will work. For that purpose to be realised, we have to be clear that the law covers all the possible circumstances—when there is a driver driving the vehicle, when the vehicle is driving the vehicle, and the circumstances between the two when somebody is handing over to the vehicle or the vehicle is handing over to the driver.
My point is that at the moment there is a gap; the Bill does not say what happens during that period. Incidentally, I do not think it matters terribly what the decision is; there just needs to be a decision, so that a case does not revolve around who the relevant insurer is under the circumstances of transition.
I know we are not debating clause 3, but since the Minister referred to it, let me point out that clause 3(2) makes it the driver’s responsibility if a vehicle is unsafely allowed to be driven automatically. A driver could be at fault if they cause an accident at the moment of transition by failing to respond when the vehicle tells them to take over, so clause 3 could actually make things worse for the driver.
Actually, I think the hon. Gentleman understates the problem with clause 3(2), which the Committee will consider in due course. During our consideration of clause 1 this morning, I made the point that unfortunately clause 3(2) contains the word “wholly”. It is therefore completely unclear what happens if an accident is not wholly due to the driver or to the vehicle, but is partly due to each, as it would be during the transition. That is a muddle, and the whole point of the Bill, which I applaud, is to avoid muddle. Muddle encourages courts to base decisions on common sense or common law, because the statutes do not tell them how to handle the circumstances. That is not what we are trying to achieve; we are trying to clarify and make certain.
We therefore need clause 2 to set out clearly the three possible situations. If the driver is driving, the driver’s insurer is liable. If the car is driving, the car’s insurer clearly has strict liability, novel though that concept is. But we need a decision—I do not really care what, so long as it is clear, definite and permanent—about what happens during periods of transition, however long they may be and under whatever circumstances they may arise. We cannot tell in advance how long the transition periods will be, and we should not take any advice from the industry that they will be only for 10 seconds and will always work perfectly—they will not.
May I welcome you to the Chair, Mr Bailey? Our discussion this morning was lively, but productive and wholesome. I am keen to make progress, as I am sure other Committee members are. The amendments tabled by the hon. Member for Eltham relate to issues that we have already addressed, but with further consideration of the transition between autonomous and human driving. Clause 3(2) states:
“The insurer or owner of an automated vehicle is not liable…to the person in charge of the vehicle where the accident that it caused was wholly due to the person’s negligence in allowing the vehicle to begin driving itself when it was not appropriate to do so.”
I am conscious that much of the debate on these amendments relates to clause 3, so I must be careful not to stray into premature consideration of a clause that the Committee has not yet reached. Nevertheless, in resisting the amendments, it is pertinent for me to refer the hon. Gentleman and my right hon. Friend the Member for West Dorset to the Road Traffic Act 1988. If the driver has some role in the accident—if the vehicle is not self-driving, either during or before the transition—the current framework, which is set out in the Act, will apply.
It is also worth saying that if a driver negligently decides to hand over control of the vehicle, clause 3 will apply, which is why I said we would end up debating clause 3 if we were not careful. If it is partly the driver’s fault, subsection (1) will apply; if it is wholly their fault, subsection (2) will apply. For example, if the driver of a vehicle designed only for self-driving on a motorway is injured after putting it into self-driving mode on a rural road, the insurer’s liability will be reduced under the contributory negligence principle. If a court finds the driver to be wholly at fault, the insurer will pay only the third parties involved in the accident. Partial responsibility is therefore addressed in the Bill and the transition, to which my right hon. Friend the Member for West Dorset paid particular attention, is dealt with in as much as we have an existing framework that of course insurers have built their current products around, which is drawn from the Road Traffic Act 1988 and other national and international regulations.
I will be happy to do that when further inspiration reaches me. In the interim, while I wait for that inspiration, I will say that we recognise the need to ensure that the transition controls are safe. It is of value to emphasise that research, including some being carried out in the UK, will help to determine a safe transition process to inform international safety standards of the kind I mentioned earlier. In essence, therefore, the field is a developing one in which those international standards are being built on. Research is taking place here and elsewhere.
The research that we spoke briefly about in the witness sessions is such that it includes the development of software to take account of endless eventualities that might occur while a vehicle is being driven or driving itself. The work being done is to simulate a range of road conditions and circumstances in which any car might find itself at any point in time on any kind of road. That is of course as numerous as might be imagined, but the aim is to have software that is clever enough to deal with all kinds of driving circumstances. The work is not complete but ongoing, and is being done on London roads as we speak—trials on London roads in real time.
I am therefore confident that the further work will lead to an outcome where the software that in the end allows us to see the further development of automated vehicles will be able to replicate circumstances that drivers find themselves in. That, by the way, relates to a debate we had earlier about the judgments that might be made by a human being replicated by the software given all kinds of different challenges.
Will the Minister focus his mind on a specific example? We are in a case in which the car has been driving itself on a motorway. It is programmed to turn off the motorway, but it is not judged by the Secretary of State to be a car of a kind that would be safe to drive off a motorway. It has therefore been programmed to hand over to the driver when it leaves the motorway—this is one of the situations on which the amendment of the hon. Member for Eltham is focused—and the driver is profoundly asleep, having been asleep all the way from London to Bristol on the motorway. The machine tries to hand over to the driver.
I am sure the Minister is right, that the software will be highly developed and it will try to hand over quickly, as far as it can, and that if it does not hand over quickly it will take all sorts of other sensible evasive action to prevent an accident occurring in such circumstances. If we could be absolutely certain that the software was perfect, we could all relax. The Minister would not need the Bill because there is no need to insure things that are absolutely perfect; they never have any accidents so there are no risks and no need for the law.
In introducing the Bill, however, the Minister rightly envisages that the software will not be perfect because things invented by human beings never are, unlike things invented by the Almighty that the Minister believes in. There will be circumstances in which the software goes wrong, such as if it tries to take evasive action having tried to hand over to a driver who was asleep and who it has failed to wake up. We have a prolonged transition period during which this magnificent software is trying and failing to get the driver to wake up and somehow does not do everything perfectly, and then there is an accident. Under clause 2(1)(a), is the vehicle driving itself in those circumstances or not? I do not know and a court will not know. It is trying not to drive itself—it is programmed not to be—but it has failed not to be driving itself. Somehow or other, that circumstance needs to be covered here. If the Minister can explain how the Road Traffic Act, which I looked at when it came up in the oral evidence sessions—
I do apologise. If the Minister can explain how the Road Traffic Act solves that problem, I am all ears.
I had forgotten for a moment that it was an intervention. Those who seek perfection on earth are invariably either extreme zealots or delusional, or both. Perfection exists only in heaven, as my right hon. Friend knows. The insurance industry does not claim that there would be no accidents in any circumstances as a result of automated vehicles, but it told us in the oral evidence sessions that it thought there would be fewer. It said that that would have an effect on the insurance marketplace because of the effect on safety—that is the exchange we enjoyed earlier—that comes about because the fallibility of men and women as drivers means that 95% of accidents, or a figure close to that, are caused by human error of one kind or another. We are clear about that.
We can also be clear that the Bill is welcomed by the industry because we were told so by Mr Howarth in the oral evidence sessions. He said:
“I think it is very clear that the legislation and broadly the development of automated driving are something that insurers are genuinely enthusiastic about.”––[Official Report, Automated and Electric Vehicles Public Bill Committee, 31 October 2017; c. 7, Q11.]
The insurance industry thinks that the Bill is an important first step, of the kind I described earlier, in establishing a framework, but it is a framework and further changes will be necessary as technology develops. Those changes will have to be dealt with in a regulation or subsequent measures.
I will give way in a moment; I just want to complete this thought. Manufacturers have spoken about creating geofenced vehicles that would operate in defined parts of the city; others have spoken about systems that would operate on motorways and other high-speed roads. It is likely that the relevant global regulations that will be used to type-approve automated vehicles will reflect such limited-use cases. It is also possible that the regulations will contain requirements that the vehicle be able to detect where it is so that the system cannot be used in other situations.
Therefore, it is not clear that we need to make matching regulatory changes in our domestic framework. If necessary, we can use existing powers—this relates to what I said earlier—in the Road Traffic Act 1988 to revise existing or create new road vehicle construction and use regulations to reinforce the global regulations. That is exactly the point that I would make to my right hon. Friend the Member for West Dorset. If that legal power exists, and as long as the Bill does not counter it—it is a useful addition, but it does not negate any of that—it seems to me entirely possible to deal with those technological changes.
I do not think that anybody could possibly be convinced by that, because it does not address the issue. The issue is when the insurer of the vehicle will be liable. It does not matter what regulations are made; they will have no impact on that question if the primary legislation says what it says now and no more. It will remain unclear what will happen in circumstances where it is not clear whether the automated vehicle is driving itself according to the terms of clause 2(1)(a), because it is in transition but failing to transition. That is a problem that the Minister cannot address through regulation; he must address it in the primary legislation if he wants the court to be clear about who is liable.
I beg to move amendment 2, in clause 3, page 3, line 4, at end insert—
“(3) The Secretary of State may by regulations define when it is and is not appropriate for a person in charge of the vehicle to allow the vehicle to drive itself.”
This amendment requires the Government to provide regulatory guidance for when it is and is not appropriate for a person to allow an automated vehicle to drive itself.
It is always a pleasure to serve under your chairmanship, Mr Bailey. I do not intend to keep the Committee terribly long on this issue. As the Bill is drafted, the
“insurer or owner of an automated vehicle is not liable”
where the event was caused by a person allowing the vehicle to drive itself
“when it was not appropriate to do so.”
The Bill does not define when it is and is not “appropriate to do so”. Our amendment requires the Government to provide regulatory guidance on when it is and is not appropriate for a person to allow an automated vehicle to drive itself.
This goes to points made previously by members of the Committee, not least the right hon. Member for West Dorset. It would clearly not be appropriate in some circumstances for vehicles to drive themselves. For example, early automated vehicles might be deemed safe to use only on motorways and not on some urban roads; or, for example, a software issue might arise such that using the automated function at that point would be absolutely inappropriate. It appears to me that the true intent of subsection 2 was to focus on bimodal vehicles, because it does not seem to apply to fully automated vehicles. Perhaps the Minister can clarify the position in his response.
One of the primary purposes of part 1 of the Bill is to provide a framework to give insurers, manufacturers and potential users greater clarity, providing confidence and encouraging progress on automated vehicles. However, it is still not clear from the Bill what the Government have in mind about when use of those vehicles would be inappropriate. I do not propose to press the amendment to a vote at this stage; I think the Minister has got the point I am making. It has been made and reiterated several times by members of the Committee. We are simply asking for regulations that better define those circumstances to be brought forward, because we cannot afford any confusion here. People must be absolutely clear where their obligations lie if we are to see the growth of the industry, which is something we all want. We do not want to leave these issues hanging over us.
I will address the points the shadow Minister has raised in a moment. Before I do, I want to come back to a fundamental point about the drafting of clause 3(2)—if you will allow me to do so now, Mr Bailey, rather than in a stand part debate—because it is relevant to the rest of the question. My concern relates to the word “wholly” in subsection (2). We discussed this point earlier today. My right hon. Friend the Minister said to me and the Committee that clause 3(2) was meant to solve the problem that I am worried about, which is that there are circumstances under which strict liability for the insurer of the machine is inappropriate, because the driver may do something either immediately before or some while before handing over to the machine that means he or she should not have handed over to the machine. Those are the very circumstances that the shadow Minister is also concerned about.
The Minister directed my attention to clause 3(2) as the solution. I pointed out then—I will now expand on the point—that if subsection (2) is intended as a solution, it is in desperate need of redrafting. The word “wholly”, which I assume has been inserted mindfully by parliamentary counsel, has a very definite meaning: it means “wholly”. Courts know perfectly well what to do with that when they come across a statute that very unusually—this is not something that we normally find—says that a contributory agency is not contributory, but absolute, and the person in question is wholly responsible. The court will interpret that very strictly, and rightly so, otherwise what on earth are we doing drafting Bills and Acts of Parliament?
There could be a circumstance under which the driver was wholly the cause of the accident. Incidentally, I cannot quite think what that might be. It is a pretty remote circumstance, and I would be interested to know whether the Minister can think of an example, but I accept the possibility of such a thing. Most of the time, however, it will be jolly tricky to work out who is actually responsible.
Let me go back to my example of leaving the motorway, but this time the driver was awake and flicked a switch that specifically made the machine take over. Let us imagine that the technology allowed that—it might or might not, we heard conflicting evidence on that, but suppose that it did—and the driver thought that the circumstances were such that the machine could take over and the machine thought, and that is probably an appropriate word to use, given that it is artificial intelligence, that it was appropriate for the machine to take over. However, they were both wrong. The machine was not good at handling the circumstance and it crashed. The machine got it wrong because it should not have taken over, and the driver got it wrong because they should not have asked the machine to take over. Who has caused the accident? I do not know. I am absolutely sure that there are people who will make millions and millions of pounds, and they are the QCs who will argue such cases in court, along with the rafts of solicitors and the enormous apparatus that goes with that. They will all be arguing about who is responsible.
If we lose the word “wholly”, we eliminate that argument, which I assume is the point of putting it in, because, as clause 3(2) is drafted, it says, “If there is the slightest doubt about whether the machine was in any scintilla of a way responsible for the crash, the driver is not wholly responsible and therefore the machine is wholly responsible, so there is strict liability for the insurer of the machine.” It may be that that is what the Minister wants to do, but it is a very odd thing to do, because the costs of insuring these machines would go up compared with what they would otherwise be. Under circumstances in which the driver was a heavy contributor to the cause of the accident by handing over inappropriately, the insurer of the machine would nevertheless be strictly liable because the machine made one millionth of the contribution to the cause of the accident. That is the effect of clause 3(2) as drafted, and I do not believe that that can be the Minister’s intention. That needs looking at.
Turning to the point made by the shadow Minister on regulations and clarification, I agree that it should be perfectly possible to handle the question of when it is appropriate or not to hand over through secondary legislation. I suspect that it will not be the kind of secondary legislation that we have been used to in the main hitherto. It will be very complicated legislation, because it may have to specify processes rather than results. I do not believe that the technology is likely to develop in a way that will make it obvious to the driver in advance, by reading some kind of guide, when the driver is meant to hand over and when not. I suspect that will be interactive and dynamic, and I suspect that the Minister’s successors—the Secretaries of State who will do such things in regulation—will have to find some way of compelling the manufacturers to create an apparatus that tells the driver in a dynamic and interactive way, as they are driving along, whether, as a matter of fact, it is safe to hand over to the machine or not.
One way in which that could happen is the way we were presented with in the evidence sessions. The machine invites the driver to take over and then there is a simple double rule: only machines that invite drivers, as opposed to giving them instructions, are allowed on the road—and, while we are at it, only those certified by the Secretary of State as being safe when they offer the chance to take over are allowed—and, moreover, the driver is never allowed to hand over to the machine except when it does offer that. That is a possible configuration. That would be quite a complicated piece of secondary legislation, because it would have to be accompanied by a series of quite complicated technical codes that ensure that it is put into practice and that the cars manufactured fulfil all those requirements.
There are of course many other models, but it is terribly important to recognise that if the Minister wants to achieve clarity here—as I think he does, and rightly so—as well as getting the drafting of clause 3(2) right, so that it is clear under what circumstances there really is liability for the insurer of the machine when there is a mixture of causation, he needs to recognise that there will need to be either a quite large superstructure of regulation that gives us clarity about the circumstances under which handover is appropriate or, at least, processes that make it unnecessary to have such clarity in a set of rules. I hope that he will recognise in his closing remarks that even if the Bill does not give new powers to do that—because he believes he has somehow got them already—he will consider all those questions anon, as well as looking at the drafting of subsection (2).
My aim is to do that a lot more quickly than you might imagine, Mr Bailey. I accept entirely that there will be a need for a regulatory framework to ensure both the safe deployment and safe use of automated vehicles. The autonomous insurance measures in the Bill are part of that, but the subsequent regulations that ensue will be part, too. They will be—necessarily—dynamic and, I suspect, quite complex, because this is a complex and evolving field. The reason that it is better done in regulations is obvious: we cannot keep bringing primary legislation to the House in such a highly dynamic set of circumstances. It is therefore absolutely right that it is done in a regulatory framework down the line.
Let me try to deal with the “wholly” issue, because it is important that we do so. If the driver is partly negligent, clause 3(1) applies, and contributory negligence would therefore also apply. Clause 3(2) is there to pick up the limited circumstances in which the driver is wholly at fault—that is, contributory negligence does not apply because it is clear that fault lies with the driver. If we did not include “wholly”, there would be a gap in the scope of the clause, as subsection (1) covers only contributory negligence. That is why the word “wholly” is in the Bill.
I am in a slightly odd position because it is the Minister’s Bill, so I would expect him to understand it better than I can, but I have to say that if that is his intent, the plain words of the text do not do the job. In clause 3(1)(b), it is perfectly clear on the face of it that the accident has to be, to some extent,
“caused by the injured party”.
That is not the circumstance we are talking about. We are talking about a circumstance in which the accident is wholly caused by some combination, but unknown, of driver—ex or to be—and machine, not by the injured party, so I do not see how clause 3(1) solves the problem of clause 3(2) having a hole in it.
Yes, but clause 3(1)(a) says that
“an insurer or vehicle owner is liable under section 2 to a person (‘the injured party’) in respect of an accident”,
so it covers both the driver or another party. That is repeated in paragraph (b). I do not understand what my right hon. Friend’s problem is.
The Minister is being very patient. Perhaps I am misunderstanding, but I beg the other members of the Committee to read the text:
“Where…an insurer or vehicle owner is liable…to…an injured party…in respect of an accident”.
The injured party is someone who has been injured—that is the reason for the reference to an “injured party”—but if I am the driver and in this case I am not injured, the insurer is not liable to me. I have just handed over control of the vehicle and it has injured somebody else, so I am not an injured party, and the injured party has not contributed to the accident, so clause 3(1)(b)—
“the accident, or the damage resulting from it, was to any extent caused by the injured party”—
does not apply. Clause 3(1) therefore does not apply in such circumstances, so it cannot solve a problem in clause 3(2) because it does not apply to the circumstances that we are talking about under clause 3(2)—or at least not to the circumstances that are worrying the Committee and that we have been talking about more or less all day, which is the question of what happens when I am handing over.
I am comfortable with the idea that the driver might be the injured party, and my right hon. Friend comfortable with that too. We are clear on the issue of whether the car was being driven by the driver or was in autonomous mode. Is my right hon. Friend concerned therefore about another party, unrelated to the vehicle, who might be affected by the accident? Is that what he is getting at? I do not understand.
I will try to make it as short as I can, but I am trying to advance the cause of understanding between us by answering the Minister’s question. We are envisaging circumstances in which a driver hands over to the vehicle and the vehicle takes over, but it turns out that it was arguably not safe or sensible for the driver to have done that. The driver was not injured and is not the injured party—the insurer is liable not to the driver, but to someone else who got damaged. That is the injured party. Clause 3(1) does not apply. That is the problem and that is the reason why clause 3(1) cannot solve the problem of clause 3(2).
I will reflect on that. It is clear to me when clause 3(1) and clause 3(2) do apply, but it is a reasonable question to ask where the clause does not apply—as my right hon. Friend has described—and what would apply in those circumstances. I am perfectly prepared to reflect and to come back with a clear answer. I am now certain to what he was referring, and that will help in the process of trying to satisfy him.
I was not able to be as short as I had hoped—I began this brief contribution by saying just how brief it would be. In respect of the shadow Minister, I think I have been clear that it is likely that the first autonomous vehicles will be used, as I said, in particular circumstances —earlier I talked about geofencing. It is likely that the global regulations that will be used to type approve autonomous vehicles will reflect those limited cases. It is therefore not yet clear that we will need to make matching regulatory changes in our domestic framework, as I have also said.
We do have the powers under the Road Traffic Act, as I said in response to an earlier intervention, to revise or create new road vehicle construction and use regulations. In that sense, the amendment would duplicate existing powers so really it is superfluous. Its intention is good, because it intends to do what I have just described, but I am not sure that for this purpose it is the right vehicle— I hesitate to use that term because, as so often in the debate so far, we are speaking about roads, journeys and vehicles. None the less, I am confident that we have enough powers and are taking enough powers, through the application of the regulations that I have said will ensue, to satisfy what the hon. Member for Kingston upon Hull East intends. On that basis, I hope that he will withdraw the amendment.