AI in healthcare: the future of health monitoring devices
By Kelly Bracken
Artificial Intelligence (AI) is of great interest not only to the legal profession but also to businesses that are affected by it. Specifically, for the healthcare sector AI presents significant opportunities both clinically and administratively. As the NHS is under constant financial strain, health monitoring devices may be of interest to relieve some of this pressure. However, the wide use of data that such devices require is still being questioned. The legal implications that arise remain uncertain as this is a new and developing area. Even though there are regulations governing the collection of health data, they may be considered too complex and unsatisfactory for many businesses. Moreover, it is not clear who would be liable if the device missed a change or misdiagnoses a patient and what area of the law should deal with this. Despite this, it is crucial that the law develops as no clear authority may lead to more uncertainty.
Social benefits and disadvantages
Patient monitoring by the NHS is amongst many administrative tasks that could easily be replaced with technology. This would enable a medical professional to spend more valuable time with patients and it will allow them to live at home rather than having long unnecessary stays in the hospital.
Because AI devices are better at sorting through vast amounts of data the health monitoring devices may be more precise in diagnosing a patient or detecting an anomaly. It has been reported that medication errors alone lead to the death of more than 22000 patients in the UK. A major cause of this is the failure to properly monitor patients.
The collection of data is of crucial importance in relation to AI devices as this ensures that the device is able to make accurate and informed decisions. Health monitoring devices are constantly collecting data , which results in acquiring a significantly larger amount of data. Information from all these patients can be put together and irregularities and other patterns may emerge. This means that it is not only important for the future treatment of that particular patient but also for other patients who may suffer from the same condition.
However, there has been a lot of debate around the collection and wide use of patients' data. Many people are sceptic of it as they are worried that their data may not be adequately protected and consequently be misused. However, as there is law governing health data collection the uncertainty is decreased. Nonetheless, complexities arise here as the NHS is underprepared on how to handle large quantities of data and how to securely transfer it to (private) third parties. This has to be done before the NHS can completely delve into the possibilities that AI has to offer.
AI and sophisticated analytics software allows large volumes of patient data to be processed, revealing patterns and correlations that might not otherwise be apparent. Data might be processed on an anonymous basis, which is suitable for large-scale research, but does not allow for diagnosis or treatment in individual cases. Where any element of individual or personalised healthcare is involved, systems must be designed and run in a way that complies with the stringent requirements of the General Data Protection Regulation (GDPR).
The law on collecting health data is largely governed by articles 6 and 9 of GDPR and s10 of the Data Protection Act 2018 (DPA 2018). Article 9 categorises health data as a special category of data, meaning that it is necessary to identify a lawful basis for processing under GDPR Article 6 and to meet the additional safeguarding measures which are set out in s10 and Part 1 of Schedule 1 of the DPA 2018. If consent is relied upon as the lawful basis for processing health data, then it must meet the stronger test of "explicit consent". This is a particularly fragile basis for processing, as consent may be withdrawn at any time. Consequently, it may be preferable for businesses to consider the potential relevance of other grounds, such as processing that is "necessary for the purposes of preventive or occupational medicine, for the assessment of the working capacity of the employee, medical diagnosis, the provision of health or social care or treatment or the management of health or social care systems and services on the basis of Union or Member State law" or pursuant to contract with a health professional and subject to duties of confidentiality.
Alternatively, businesses may take the view that their processing of health data does not engage GDPR because the data is fully anonymized. Any such decision requires considerable care. If data is not anonymized, but is only "pseudonymised", then GDPR will apply. In those circumstances, the strength of security measures applied to prevent the reversing of "pseudonymisation", so that data can be individually identifiable, becomes particularly important.
A barrier to innovation?
AI is a developing area making the legal uncertainties in relation to liability enormous. Even though it is often thought that these new forms of AI are more reliable, and may reduce the deaths relating to medical malpractice, there remains an element of unpredictability because they are machines.
It is important that the law takes an approach to this because deciding these issues on an ad hoc basis may result in even more uncertainty and unpredictability. This may pose not only an issue for the NHS itself but also the companies that have manufactured the products and the businesses behind the idea.
The legal complexities that arise around AI make it difficult to apply the law which is applicable to liability for medical practitioners. The question of who should be accountable has multiple options including:
- The insurer
- The product manufacturer
- The user (hospital or medical practitioner)
- AI itself.
The answer may well be found within an existing area of the law or a combination of various parts of the law. The areas that may be of interest in relation to health monitoring devices are the law of negligence, vicarious liability and consumer protection. If the product in itself was faulty then liability may be easily placed upon the manufacturer under consumer protection law. However, it is more complicated when there is a misdiagnosis which relates to the defect of not the actual product but the software. Here the healthcare provider may be held vicariously liable for the AI’s negligence if this could be established in relation to the monitoring device’s action. Lastly, AI could be given a legal personality which would mean that it could be accountable for its actions. The hospital, as a consequence, could be held vicariously liable for AI’s actions.
Due to the fast-developing nature of AI, the law is struggling to keep up with these changes. If no action is taken soon to clear some of these ambiguities the courts will be faced with a serious challenge to resolve these matters when they come to court. As we become more dependent on AI, this realisation may come quicker than we think.
 Alex Matthews-King, ‘NHS medication errors contribute to as many as 22000 deaths a year, major report shows’ Independent (London, 23 February 2018) https://www.independent.co.uk/news/health/nhs-medication-errors-deaths-prescription-drugs-jeremy-hunt-york-university-health-a8224226.html accessed 8 March 2019
Shipping container shops
By Lily McDermaid
Shipping containers are used worldwide for transporting all sorts of cargo, but they are also readily adapted for use as low-cost accommodation. Therefore, using old containers to house people and businesses seems wonderful. There might be some unforeseen and jurisdiction-specific legal issues, so businesses should seek expert advice before investing. For example, it might not be easy to obtain suitable and affordable insurance cover for shipping container homes/businesses. Further, the bureaucracy involved with planning permission and building codes might put more obstructions in the way, making this sort of development difficult for all.
Starbucks has built shipping container complexes in Taiwan and Seattle. Container City is a home and business complex in Brixton, London. However, as more containers can always be added and more businesses and tenants move in, there could never be joint tenancies (Antoniades v Villiers  UKHL 8). This opens up the possibility of occupiers only holding a personal licence to be there. The obvious implications of this are that these licensees would have less protection against unfair rent and eviction. It is important not to forget about an issue like this even with a new form of construction like shipping container homes. It could depend on the layout of the complex as to whether you could validly treat each container as a separate property or whether each room is so reliant on the shared spaces that they are like separate rooms in a house. This difference could prove important to inhabitants. As this relies on the specific layout, surveyors and the Land Registry could be left with quite a headache if these container complexes become more widespread and more popular.
There is also the question of how insurance would work on a large scale. For instance, should each container be insured separately? At least this way if there is damage or a theft during construction of the whole complex then it might be easier for individuals to recover specific losses. This would result in lots of extra paperwork for the site manager or the tenants though, costing more time and money.
Should a container complex be insured as a whole? This might require valuations every time that a part is added or taken away from the ‘whole’ as well as difficulties with insurance payouts and covering premiums. It might however be more beneficial in complexes where there are lots of shared, integrated spaces.
Does it depend on how connected the container is to the land? If so, barely connected containers that are easily removable may be insured similarly to a caravan. Then, the issue arises of what is part of the land. The most influential case law in this area is Elitestone v Morris  2 All ER 513, which concerned the legal status of a wooden bungalow. If it were classed as a fixture, the occupier would have been protected from unfair eviction by the Rent Act 1977. The House of Lords instead categorised this bungalow as a chattel, merely placed upon the land, meaning that the Rent Act protection did not apply and the occupier could be removed. Applying this to shipping containers, the courts could potentially follow the decision in Elitestone, leaving occupiers in an insecure position. Elitestone, in application, means that to have a protected tenancy, there must be a legal basis for occupying the land, like a lease or contractual licence.
Arguably, this means that there is a need for legislation to clarify the law on altered shipping containers and the protections afforded by different legal means. The only recent piece of legislation anywhere near this is the Mobile Homes Act 2013, which is simply a set of amendments to an earlier Act that only applies to caravan sites. Therefore a shipping container may come under the definition of a caravan (Caravan Sites and Control of Development Act 1960), changed to mobile home in 2013 (Mobile Homes Act 2013 and Mobile Homes (Wales) Act 2013). However, this simply is not clear enough. For instance, these Acts apply to structures designed or adapted for habitation only if it is capable of being moved from one place to another. It does not apply to those made of 2 or less sections to be assembled on site as it would not be legal to transport it on a highway, with some restrictions on dimensions. This raises more questions than it answers. What happens if a container rusts to the point where it cannot be moved? What if two containers were adapted into one home? At what point does a shipping container go from being defined as a caravan to being a permanent structure? Illogical distinctions exist that could have easily been dealt with in the 2013 legislation. Therefore, it is down to Parliament to reform this niche and technical area, before it potentially becomes more popular and more difficult to regulate.
Plain old shipping containers go for around £1,500 on eBay. Renovations can cost from around £7,500 to £125,000, depending on how extravagant a lifestyle the occupant wants. Seeing as the shipping container itself needs to be strengthened, made watertight and protected against rust, if this is not done properly or by amateurs, then it might be quite difficult to get insurance. Getting insurance for a non-standard property is not easy anyway, as it takes longer to insure constructions that do not meet the general criteria for a house. This requires inspections, which, depending on where it is, relies on strict adherence to building codes. In cases where amateurs have made their own permanent, standalone house, there has been issues with them not even knowing that building codes exist let alone how they should have adhered to them. This could cause long-term consequences for individuals or small businesses, putting them at risk of financial loss if the worst were to happen and their new home or office be destroyed or damaged.
The decision to live in a container home is not a stable one. Eviction proceedings were brought against a man in Oxford in 2012 when his local council deciding the shipping container he was living in was unfit, meaning he instead faced living in his van.
It could prove quite difficult for this to get off the ground because of extra hoops owners have to jump through. For this reason, it seems like this new architecture trend may not change the UK just yet. If it did, or if it became popular in a different country, the housing market could be devalued slightly, making it easier for people to get on the housing ladder and start their own businesses. This would affect mortgage providers as mortgages would be for lesser values and the capital property would fetch less if sold upon repossession. It could, however, spark demand for a new specialism in insurance, providing scope for other industries to profit from the potential change.
"Fraud unravels all", or does it?
By Victoria Zubareva
Arbitration awards often have to be enforced in the court of a foreign state where the defendant’s assets are situated. Under s 103 of the Arbitration Act 1996 an English court has a number of grounds for refusing to do so, for example, if such enforcement would be contrary to public policy. One of the instances when public policy might be engaged is where there is some kind of fraud involved in concluding or performing the underlying contract. However, public policy often meets with another important principle of arbitration law – finality of arbitration awards. These principles were considered in Sinocore International Co Ltd v RBRG Trading (UK) Ltd  EWCA Civ 838, where the Court of Appeal enforced a Chinese arbitration award in spite of alleged fraud involved during the performance of the underlying contract.
Under a sale contract Sinocore was to ship rolled steel coils and RBRG was to make payment by a letter of credit that it opened in a Dutch bank. The parties amended the contract to require RBRG to arrange a quality and quantity inspection before or during loading. Then, instructed by RBRG, the bank issued an amendment to the letter of credit, that had not been agreed to by Sinocore. The required shipment period was changed from “July 2010” to “20th to 30th July 2010”. After the goods had been loaded, bills of lading dated 6 July were issued and Sinocore sent a shipping notice to RBRG the same day. On 22 July Sinocore requested payment under the letter of credit presenting bills of lading dated 20–21 July 2010. It was common ground that these bills of lading had been forged to match the changed date in the letter of credit. The bank refused to pay following the injunction obtained by RBRG in a Dutch court. Sinocore commenced court proceeding against the bank in China but its claim was dismissed on the ground that the bills of lading were forged.
In the China International Economic and Trade Arbitration Commission (CIETAC) arbitration proceedings the tribunal awarded damages to Sinocore for breach of contract by RBRG. RBRG made an application for setting aside the order of the High Court to enforce the award, claiming that it gives effect to a claim by Sinocore that is based on forged bills of lading. The application was dismissed and the Court of Appeal upheld the decision of the lower court for several reasons.
Finality above public policy
It was emphasised that courts should give a restrictive interpretation to public policy. In particular, if the tribunal, staying within its jurisdiction, has ruled that there was no illegality involved under applicable law of the contract, the court should not go behind the award to re-investigate this issue unless “possibly in exceptional circumstances”. Thus, the principle that the finding of the tribunal should be final outweighed public policy against illegality.
Causal link required
To refuse the enforcement of an award on the grounds of public policy, illegality has to be causally linked to the claim on the basis of which the award was made. For example, while awards where the contract was alleged to have been procured by bribery have been enforced (National Iranian Oil Co v Crescent Petroleum Co International Ltd  EWHC 510 (Comm)), an award giving effect to a contract to bribe would not be enforced. In Sinocore v RBRG the courts accepted the tribunal’s findings that there was no causal link present. The arbitrators held that the cause of the termination of the contract and Sinocore’s failure to obtain payment had been RBRG’s instructing the bank to issue a non-conforming amendment to the letter of credit without Sinocore’s consent. Thus, forged bills of lading had been presented after RBRG had incurred liability for breach of contract.
Rethinking the maxim: unified approach
There is a long-established (Holman v Johnson (1775) 1 Cowp 341, 98 ER 1120, Lord Mansfield) rule of English public policy that courts will not be used as enforcers of fraud. It is based on a maxim that once committed, fraud unravels all, that is, taints the whole contract and the claim based on it. Dealing with RBRG’s argument to this regard, Hamblen LJ pointed out that as neither RBRG, nor the bank had been deceived and Sinocore had not benefited from their act, that was “at most a case of attempted fraud”. Given the tribunal’s ruling on causation, this attempt was only collateral so that in enforcing the award the court was not allowing a dishonest person to use its procedures to enforce a fraud.
Interestingly, this reasoning of the Court of Appeal is in line with the latest trends in another field of law, namely, insurance law, where the same maxim found its application. Until 2016 a claim made by the insured using fraudulent means and devices was considered to be a fraudulent claim. That was a genuine claim, based on a real loss, in support of which the insured provided the insurer with false statements or false evidence, for instance, a forged invoice that had been used to obtain funding from a finance company (Sharon's Bakery (Europe) Ltd v Axa Insurance & Anor  EWHC 210 (Comm)). The remedy for fraudulent claims, codified in s 12 of the Insurance Act 2015, is that the insurer is not liable to pay the indemnity and can also terminate the policy as from the date of the falsehood. However, recently in Versloot Dredging BV v HDI Gerling Industrie Versicherung AG  UKSC 45 the Supreme Court held that such subsequent conduct (“collateral lies”) in making the claim did not affect the right of the insured to the indemnity if he had suffered a genuine loss.
All in all, the approach of the English courts towards attempted document fraud allows those who have genuine claims that arose prior to falsehood to recover what they are really entitled to. In the context of arbitration, it also helps to make England an attractive forum for enforcing foreign awards.