AI for Banks: Key Ethical and Security Risks
The use of AI to drive efficiency and business improvement has increased significantly in the last five years across multiple sectors.
Technological innovation continues to disrupt the status quo in established industries, presenting both opportunities and threats to established industry leaders. In this article, we focus on the disputes that might arise from changes in the energy sector, and emerging avenues for avoiding and resolving such disputes.
We are in the midst of the “fourth industrial revolution”. As discussed in our guide Unlocking the blockchain: Digitizing the energy value chain, distributed ledger technology (DLT), and the applications enabled by it, has the potential to lead to a fundamental change in how the energy industry shares data and transacts, and may even alter market structures.
|Lost in the lingo? Speed read our Legal technology jargon buster|
DLT refers to software applications that deploy a digital database of transactions (i.e. a ledger) which is distributed (i.e. identical copies of the ledger are maintained on multiple computer systems). A common iteration of DLT is “blockchain”. The significance of DLT is that it allows data gathering and record-keeping on enormous scales and facilitates the efficient sharing of certain data between parties to a transaction. DLT has important B2B applications throughout the energy sector, and is potentially disruptive at every stage of the value chain including in asset management, trading, commodities tracking and certification, transport and logistics and payment mechanisms.
When combined with data analytic software, DLT is also an incredibly powerful B2C tool. It enables much deeper understanding of customer attitudes and behaviors and closer monitoring of market trends, facilitating closer interaction with retail customers.
DLT also has the potential to automate many activities, when used in combination with smart contracts. A smart contract is a “set of promises, specified in digital form, including protocols within which the parties perform on these promises” (Nick Szabo, Smart Contracts: Building Blocks for Digital Markets, 1996).
A smart contract is
As smart contracts are simply automating code other technology could also be used, but smart contracts are typically deployed on DLT.
A key virtue of a smart contract is its “disintermediation”, in that the parties can transact on a peer-to-peer basis, without the need for an intermediary – whether that be a central authority or third party. The system facilitates correspondence between the parties’ respective copies of the ledger, and the smart contract performs automatically, as coded, when pre-agreed events or values are recorded on the ledger.
|For an in-depth discussion of smart contracts, read our guide Smart Contacts: coding the fine print, a legal and regulatory guide|
This has significant implications in the energy sector. In energy trading, for example, cost reductions in deal execution and decreased credit risk may be achieved by enhanced information flow and disintermediation, using smart contracts to automate the execution of trades.
The potential uses are even more exciting when combined with the Internet of Things (IOT). IOT refers to the connectivity between intelligent sensors and devices via a network. Sensors are already used throughout the energy industry, for example, monitoring and maintaining temperatures during oil extraction and in maximizing gas distribution by managing pressures and identifying leakages. Combining this with DLT, smart contracts and IOT could lead to unprecedented automation of operations across the industry. For example, electricity network faults could be identified using smart sensors (rather than visual inspection) and maintenance automatically scheduled by a selfexecuting smart contract.
|There are many other use cases in the energy sector for these emerging technologies – for more information read our guide Unlocking the blockchain: Digitizing the energy value chain|
However, any business innovation, whilst creating opportunities, also creates new areas for disputes. Given that much of this technology, and the potential use cases, are rapidly evolving, the exact nature of disputes likely to arise is difficult to predict. To complicate matters, most laws that will apply to such disputes were developed in a world that did not even contemplate such technologies.
One fundamental issue is whether smart contracts in fact have any legally binding contractual effect at all. Under the common law a contract requires: offer and acceptance, consideration, intention to create legal relations and certainty of terms. The coding of a smart contract may lack these essential legal elements. Furthermore, by its very nature a smart contract may not satisfy prescribed formalities for legal validity e.g. a certain form and/or method of execution. There is a clear risk to parties if commercial arrangements lack legally binding effect.
|For a more detailed discussion of whether smart contracts are legally binding, read our white paper Can smart contracts be legally binding contracts?|
Coding errors may also give rise to legal uncertainties, and these matters will be more complicated when things go wrong in situations without any human involvement. In some circumstances, it may be difficult in accordance with established legal thinking to establish liability and/or even identify the defendant against whom to seek redress. For example, if an automatically generated invoice is sent to and automatically paid by a customer but an audit later proves it to have been wrong (based on an inaccurate reading from a defective sensor or due to some small programming error in the algorithm). Another foreseeable example is where a bug or error in a smart contract results in a mistake in what the parties thought they had agreed or some entirely unintended outcome. Most jurisdictions have legal mechanisms to deal with the consequences of or correcting mistakes in contracts, but we are yet to see how the courts will apply such laws to smart contracts.
Also fundamental are questions of jurisdiction and governing law. The energy industry is global, and transactions and supply-chains commonly have cross-border elements. That brings with it questions of which court(s) has jurisdiction to hear disputes and which law(s) should be applied. Pity the court that has to determine governing law and jurisdiction in this hypothetical DLT use case example: disputes arise in respect of a DLT based commodity trading platform created and provided by a Swiss technology company, on which a South African buyer and a US seller transact, with the commodity to be delivered in Italy. It is not that there is necessarily a legal vacuum to decide these issues, and courts are of course accustomed to dealing with difficult jurisdictional and conflict of law issues. But we are yet to see how these issues will be dealt with by the courts in practice.
|For a deep dive into legal and regulatory risks, read our guides Smart Contacts: coding the fine print, a legal and regulatory guide and Unlocking the Blockchain – Blockchain disputes: Risks and resolutions|
Legal and regulatory risk in the energy sector is already a complex matrix, complicated by the fact that the sector involves global players but there is no single, universal, legal system. There is a plethora of different and often conflicting systems across the globe, and as disputes arise, multiple courts will grapple with and laws will evolve (or be legislated) to address challenges created by fast emerging new business realities. These few examples are intended only as an introduction to the novel legal risk profile of innovative technologies. The biggest risks are of course likely still unknown.
Parties can take steps to minimise risk. This requires careful analysis of the risk profile of the particular use case in the specific context. Critically, this should be done at the outset. Some issues may be ameliorated, at least to some degree, by incorporating contractual provisions in the smart contract (directly or by reference to terms and conditions). To deal with liability, for example, parties may include provisions to allocate risk or to limit or exclude liability (on a back-toback basis, as appropriate). Parties may also include provisions to deal with events that amount to force majeure or arise from third-party (mis)conduct. Jurisdictional problems may be assisted by choice of law or jurisdiction clauses.
However, the very nature of emerging technologies means that not all risks can be anticipated. Given the commercial and legal uncertainties, arguably the most important decision will be the choice of governing law and dispute resolution mechanism in the smart contract.
As discussed in our article Arbitrating smart contract disputes, arbitration is uniquely well-suited to deal with disputes relating to innovative technologies arising in the energy sector.
Arbitration often has significant advantages in respect of enforcing the parties’ agreement to arbitrate, as well as any arbitral award subsequently rendered. It offers a neutral forum, and allows parties to avoid recourse to courts where, for example, there might be issues with bias, corruption, lack of expertise, or excessive cost or delay. Importantly, parties to arbitration can also choose their arbitrator(s) – a significant benefit where the alternative would be judges or juries with no relevant expertise. As arbitration is a consensual, contractual process, parties also have greater flexibility to tailor the arbitral procedure to their specific needs.
As a process, arbitration is more readily able to adapt to the new status quo brought about by innovation than many national courts, and can offer innovative dispute resolution solutions. Innovators in arbitration are already developing automated dispute resolution processes that operate in the same way that a smart contract does.
For example, parties could agree and code into the smart contract that an arbitrator(s) will be automatically appointed from a pre-agreed list (perhaps chosen at random) and that the arbitrator’s decision is to be fed back into the smart contract giving it immediate effect and circumventing enforcement issues. This idea of “decentralised” arbitration was mooted in 2016 by Vitalik Buterin, the individual behind the blockchain platform Ethereum. This concept could be expanded to add provisions as to how the disputes will be resolved, such as a fast-track process, decided by reference to only certain preagreed external data and in accordance with certain laws or rules. For certain types of disputes, a largely automated (and cost effective) arrangement such as this would be welcomed. Other examples of innovative arbitration are set out in the table below.
At the furthest end of the spectrum is so-called “AI Arbitrators”. As business activity in the energy sector becomes increasingly automated, and ever greater volumes of data are generated and inter-connected, it is conceivable that arbitration could eventually offer wholly automated dispute resolution solutions by applying analytical tools and artificial intelligence to decide disputes based on data input automatically by the smart contract and which are automatically given effect in the smart contract. In this way technological innovation has great potential to disrupt not only the energy sector, but also dispute resolution itself.
These are exciting times for the energy sector, with emerging technologies offering incredible opportunities. Parties do need to ensure, however, that they also consider the potential new areas of risk, and ensure that they have incorporated appropriate contractual protections and an effective and enforceable mechanism for enforcing the contract and resolving disputes.
With special thanks to Khawaja Akbar, trainee, for his assistance with this article.
|“Multisig” mechanism||Enables parties collectively to nominate an arbitrator, who is empowered to decide disputes and transfer assets or money on the blockchain to give effect to the decision.|
|Bitrated||An adjudication platform for Bitcoin. Parties nominate, ahead of time, a trust agent to act as an arbiter for their transaction. “Arbiters” build an online profile containing, for example, descriptions of their approach to procedural and substantive rules. They have the power to transfer currency directly to the successful party to the dispute. The platform ranks adjudicators based on “trustworthiness”. Interestingly, those with the highest ratings tend not to refer to national law in their methodology for filling in gaps in contractual arrangements.|
|Confideal||A platform for transacting via smart contracts on the Ethereum blockchain which contains an integrated arbitration service. Allows parties to select qualified arbitrators, who resolve disputes without having access to the personal data of the parties – all they know is the parties’ cryptocurrency wallet addresses. Arbitrators’ decisions are intended to be legally binding in national courts but arbitrators do not have the power to enforce decisions themselves. Arbitrators create accounts on the platform and enter personal data: languages, location, specialization, working experience, etc. An arbitrator for any particular deal is selected by both sides of the contract during its setup, before signing. Arbitrators are publicly rated and ranked based on objective and subjective criteria. There is also a separate selective measure based on user votes for arbitrators using “likes” and “dislikes”. Arbitrators are paid up to 10% of the smart contract’s sum. There is also a mediation option.|
|Blockchain Arbitration Rules||Arbitral rules based on UNCITRAL Arbitration Rules. Under the rules, referral to arbitration triggers a function that pauses execution of the smart contract. An appointing authority performs administrative work in connection with the arbitration including nominating the arbitrator. All relevant data is contained on the blockchain, including details of the contract and email communications. Once the arbitrator has the necessary data, including statements of claim and defence, they may call an oral hearing (in person or via video conference). Depending on the arbitrator’s award, the appointing authority either resumes the smart contract, modifies its execution, or ends it permanently. The arbitrator is paid with funds available in the smart contract in dispute.|
|Datarella||A blockchain solutions provider. In July 2017, Datarella conducted an arbitration using blockchain technology, where the smart contract governing the relationship between the parties contained an arbitration clause referencing to the Blockchain Arbitration Rules.|
|Kleros Court||A decentralized application built on top of Ethereum, which works as a decentralized third party to arbitrate contractual disputes. It relies on game theory incentives to ensure “jurors” (arbitrators) decide cases correctly. Parties can choose Kleros as their adjudication protocol in the smart contract, and relevant data is securely sent to Kleros. A tribunal is drawn from the crowd, jurors evaluate evidence and “cast their vote”/render an award, which is enforced by the smart contract. Jurors collect fees for deciding disputes in the form of tokens – after the dispute is decided, jurors whose vote is “not coherent with the group” lose tokens, which are transferred instead to “coherent” jurors. The creators postulate that this will incentivise “honest rulings”.|
|Jincor Arbitrage||Jincor is a smart contract and cryptocurrency platform. Offers an “arbitrage” system. Arbitrators are chosen by mathematical algorithms with regard to “digital reputation, sphere of competence and practical experience in specific jurisdiction and economic sector”. Arbitrators’ fees are defined in advance and paid by the losing party at the end of proceedings. Proceedings are held in “an anonymous digital room”. Arbitrators remain unknown to the disputing parties. The losing party is given “a commitment requirement” and it seems that failure to comply with that (it is not clear from Jincor’s description) will result in “losing reputation and business relations, which are associated with a digital ID of the participant”. All participants of the ecosystem have access to a unified database, which will contain data on all previous arbitration proceedings. This allows parties to assess each other’s reliability before signing a contract.|
|EOS Core Arbitration Forum (ECAF)||The dispute resolution mechanism for resolving disputes between parties operating on the EOS blockchain, a “governed” blockchain (i.e. a rules-based environment, pursuant to the EOC Constitution). Arbitration is conducted under ECAF’s Rules for dispute resolution.|
|SAMBA (Smart Arbitration and Mediation Blockchain Application)||An application created by a Miami-based start-up which aims to use DLT to offer a private, decentralized application to facilitate cost and time efficient resolution of disputes. Users will be able to input their arbitration clause or agreement into a request form, which will translate the information into code and generate the smart contract. The smart contract will then be sent to the SAMBA account of the designated arbitral institution, which must agree to administer the case, and then to the respondent so the arbitral process can begin. The final award will be stored on the blockchain. All participants (claimant, respondent, arbitrators and institution) have keys to access data at appropriate times in the process, and there is a drop box for electronic discovery. Currently a secure website but intended eventually to be an application that can be used on any device. Also looking at how arbitral institutions need to adapt to the digital era, in particular DLT.|
|DAMN (Decentralized Arbitration and Mediation Network)||A proposed network which “would be built on top of the New York Convention legal structure”. It would provide users with layers of choices regarding whether a dispute will be resolved by a person, an algorithm, pools of random jurors, through the collaboration of the parties involved or a decentralized autonomous organizations. The creators anticipate “hundreds of dispute resolution systems of different levels of complexity catering to anyone who wants to run them.”|
|Cryptonomica||An online and offline identity verification service and database of verified identities with keys for signing electronic documents, blockchain transactions etc. Cyprtonomica Ltd (formerly known as the International Arbitration and Cryptography Centre Limited (ICACC)) purports to be a permanent international arbitration authority, registered in the UK. Offers an online arbitration procedure under its own arbitration rules.|
The use of AI to drive efficiency and business improvement has increased significantly in the last five years across multiple sectors.
© Norton Rose Fulbright LLP 2021