Automated Vehicles Bill [HL]

Baroness Bowles of Berkhamsted Excerpts

2nd reading
Tuesday 28th November 2023 Lords Chamber

[Baroness Bowles of Berkhamsted (LD)]

My Lords, I declare my interest, as a member of my family works in the vehicle connectivity sector, but I have no financial interests. I have migrated to this Bill from my usual territory of financial services, business and intellectual property, and it is those aspects which I will mainly cover.

At present, there is a thriving ecosystem in the automated and connected vehicle sectors, including small and innovative businesses, especially on the software side, and whose protection and income streams are based on intellectual property and data. While I understand and appreciate the work that has been done by the Law Commissions and the legal framework in the Bill, I have questions about its impact on that ecosystem for small businesses in particular, and for the preservation of intellectual property rights and rights to monetise data.

My concern is heightened by the already prevalent tendency to default to big-name providers, including large overseas technology companies winning government contracts. There is already an unlevel playing field caused by “play safe” selection of dominant companies, and that dominance is further reinforced by the acquisition of data for themselves when they are selected for important UK projects. Government has a role to play here if it wants to emerge with any UK champions. That includes government having the expertise to analyse bids from less obvious sources, and without forcing commercial disclosures of a type not normally sought from large entities. I have also noted the announcement that the Government may be amending the National Security and Investment Act regarding mergers and acquisitions, which may also further benefit big-name companies from abroad.

Turning to the Bill, I wonder how the present ecosystem will be affected by the conditions surrounding the new regulated entities. This is a point where information requirements could force disclosures or distort contracts, essentially stripping supply chain companies of their commercial, IP and data exploitation opportunities. It would be unfortunate for smaller companies to be forced into a choice between ultimate liability and massive insurance costs, or contract and licensing provisions that leverage away their own data exploitation possibilities.

I now turn more specifically to insurance. At the commercial level, the London market is well placed to develop cover for the UK and elsewhere, so I noted with interest that the Government announced a consultation on captives: that is, an insurance company which is a wholly owned subsidiary providing insurance to its non-insurance parent company or companies. Somewhat ironically, London has the best experts for this underwriting, but we do not have any captives within the UK. Captives might be attractive for the automated vehicle sector, with their associated commercial considerations, including as a part of providing “good financial standing”. However, will the UK regulator, the Prudential Regulation Authority, take exception to the monoline nature of captives and make life difficult? Has the Minister any information in that regard, and have there been cross-departmental discussions?

The other insurance aspect is for the consumer, with a particular point of interest being when a driver becomes liable after a transition from automated driving. Naturally, insurance companies are interested in data on those transitions, including how many times those instances happen and in what circumstances and where, as well as driving patterns and so on. That is commercially useful to insurance companies for the purposes of assessing risk, getting the prices right and ascertaining whether someone is liable, and who. My question is: should they get it all for free? Some insurance companies are already involved in funding developments; others are not. Should their positions be the same? Over time, information will derive naturally from actual claims, but, more generally, this is an area where the entitlement to benefit from or monetise data resides elsewhere.

Some of this kind of information may also be delved into for other investigatory purposes, including authorisation and licensing. For the consumer, there is the dangling threat of insurance companies having to analyse whether the response at handover was what a competent driver would do.

Putting a simple time limit for the liability switch is making light of a difficult situation. Fighter pilots are trained in handovers—and I doubt that they are doing crosswords in the meanwhile; but that is the position envisaged in the documents I have read. I have spent a lot of time being driven around, rather than driving, as a safety precaution. Many times, I have looked up from reading a brief to wonder where on earth I was and felt quite disoriented—and that was in a constituency I was supposed to know and was representing. Regrettably, we will need statistics before we know the interval after handover that is fair to compare with a competent driver in ordinary circumstances—even if we can define that—because it will be different on a straight road with no manoeuvring expected, compared with a much more complex layout or circumstances when the AV cannot cope any more. Potentially, this puts a lot of uncertainty on the consumer, as well as insurance companies, and there may be a role for a levy and compensation model around the transitions, at least initially.

Clause 14 allows access to data for insurance purposes, but it is not clear to me whether that is intended to be free, so perhaps the Minister could shed some light on that. Already, there is lobbying to make the wording mandatory. That is relevant because provisions in the Bill acknowledge commercial rights—in this context, Clause 42(7), which applies to the whole of Part 1 —but that does not protect commercial rights where the disclosure and use by third parties is considered necessary. Clause 14 already seems to imply that information disclosure to insurers and others may be necessary—regulations might make it so—and the industry lobby wants to make it so in the Bill. Either route to mandating puts it outside the commercial harm provision in subsection (7). That needs more examination, and I suggest something along the lines of fair and reasonable compensation for the commercial use of information.

Finally, I have some interest in proof of safety and proof of concept, as it takes a vast amount of data to prove statistically that something is safer than something that exists already. Road fatality in the UK is 5 per billion miles driven. If you were to compare that with, say, a statistically significant number, you might have to look at the number of driven miles per 50 deaths. However, that would be 10 billion miles driven, which could be 100,000 AVs driving 10,000 miles, or 10,000 AVs driving 100,000 miles. That is an awful lot of miles, and we only have about 100 to 500 vehicles that could do that, so that converts to 20 to 100 years, depending on which end we are at.

I am interested to know how we are going to get this kind of mileage done. Will it be all in the UK? Will we accept evidence from other countries? Will it be for each AV, because why should you have your statistics marred by a bad apple? If we are going to accept evidence from other countries, can we be sure that it will be from similar driving conditions? A lot more could be discussed around that, but it is a very big statistical job to prove that something is safer than something else.

I recognise that the Bill is the start of constructing a framework and not the end, so I look forward to exploring its effects and some of the things that I have raised further at Committee stage.