Feeds:
Posts
Comments

Musing about AI and responsibility

A recent article on the Ethics of AI led me to think about responsibility for the consequences of decisions by an AI entity.

Arguably, however autonomous an AI entity, its manufacture, programming, and energising must derive, directly or indirectly from an human originator or human enterprise – a creator – accessible in a way our Creator is not – at least in this world.

That creator may have delegated duties to an operator such as the driver/pilot of an autonomous vehicle. My point is that, at some level, the most autonomous device always will be directly or indirectly traceable to a legal person; a second – or higher – level device ultimately will be so traceable.

Even when the intellect of the artefact becomes superior to that of the human originator, as it may, the artefact will remain an artefact. Its decisions are the responsibility – be it never so indirect – of a human.

There will be a question as to the human succession when the first originator dies but the law, though differing from jurisdiction to jurisdiction, can deal with such problems.

An autonomous vehicle at large is, I imagine, no different from a tiger. The owner is responsible. The English case of Rylands v Fletcher [1868] UKHL 1, LR 3 HL 330, (1868) LR 3 HL 330 provided strict liability for the release of dangerous things that cause damage. I would expect other jurisdictions to have such a law. I would suggest that it is a logical part of Natural Law.

I haven’t seen much discussion of ultimate liability in the generalisations about the ethics of quasi-autonomous entities, but I may not have been looking. I would be interested in any reaction to my proposition. There is something finite about an accessible, worldly creator.

Advertisements

Thoughts on the logic of Brexit

May I suggest that it is arguable that to adopt the result of a referendum as definitive and final is counter to the principles of a parliamentary democracy?   That is the logical implication of the judgment of the majority of the Supreme Court in the matter of Miller  ([2017] WLR(D) 53, [2017] UKSC 5).

Leaving to one side the question of whether or not the Referendum was intended to be binding – and, indeed whether voters were made aware of that, the words used in the Referendum Question being “Should the United Kingdom remain a member of the European Union or leave the European Union?”[my emphasis] – the Court made it clear that Parliament is paramount.

The OED defines democracy as “Government by the people”.  The people, not merely a majority but all of them.  That is why we delegate the task of ruling us to our choice of Members of Parliament.  It is not simply an oratorical flourish when an elected Member proclaims that she or he will represent all the constituency, those who voted for and those who voted against, to say nothing of those who did not vote at all.

The duty of a Member of Parliament was described in the great speech given by Edmund Burke to the electors of Bristol, in which he distinguished representation from delegation. The MP represents the interests of his constituents, not their opinions.  That is what he or she was elected to do, albeit on the basis of his or her declared political opinions.

In the Referendum on the European Union, the People have spoken.  Perhaps, but roughly half of them have said one thing and half the other.  Let it be clear, the decision has been the decision of the Government and now the House of Commons, not the People.  Seemingly, that is what the Supreme Court has prescribed.

At least we can be grateful that there is an Article 50.  In similar circumstances, in the 1860s, a bitter civil war in the United States of America, the American Union if you will, compelled the Confederate States to remain by force.  We are more enlightened now.  Tomorrow was another day.

‘Detail’ in Adjudication

In Amey Wye Valley Ltd v The County of Herefordshire District Council [2016] EWHC 2368 (TCC) the learned Judge observed, “Errors of fine detail are part of the process effectively accepted by Parliament as a consequence of the process of adjudication. The `right’ answer is secondary to the parties having a rapid answer.”.

The adjudicator had made a mistake in his spreadsheet.  Amey said it was 2.5 million, HDC 1.9 million.

That’s fine detail?

Arbitration – a law unto itself?

Arbitration – a Law unto itself?  Fortunately, there is no copyright in titles, as I would like to take as a text the title of the 30th Annual Lecture organised by The School of International Arbit…

Source: Arbitration – a law unto itself?

Arbitration – a law unto itself?

Arbitration – a Law unto itself?  Fortunately, there is no copyright in titles, as I would like to take as a text the title of the 30th Annual Lecture organised by The School of International Arbitration and Freshfields Bruckhaus Deringer and given by the Right Hon the Lord Mance, a distinguished Justice of The Supreme Court, on 4 November 2015.

IMG_0007 sized

Figure 1 GBH

Most respectfully, and as an Engineer, not a lawyer, I mean the phrase sincerely, I found myself in agreement with the proposition of the very first paragraph which I set out here for reference:

Lord Mance

Figure 2 Lord Mance

“This article argues that theses advocating an independent or transnational system of arbitration lack coherence. Arbitration is not, and should not become, a law unto itself. Arbitration already faces problems in maintaining coherence in its jurisprudence and confidence in its efficacy as a dispute-resolution mechanism, particularly given that no general means exist to ensure that awards are consistent. These problems could only be exacerbated by a declaration of unilateral independence.” [The emphasis is mine – GBH.]

My own, layman’s, analysis does not advocate that arbitration should be a law unto itself or indeed a law at all.  A private agreement enables one or more private persons, privately chosen without the engagement of the trappings of any legal system, to obtain for themselves and on behalf of the disputants the fair resolution of the dispute before them.  That resolution, in the form of the Award is, I argue, no more and no less than an autonomous document, recording a private decision, without any legal coercive force.

It is usual for legal systems worldwide to have respect for the maxim pacta sunt servanda.  That expression is more than a ukase; it is a statement of fact, a logical identity.  Pacts are to be served.  If words are not to be served they are words in the air but no agreement.  When persons agree, they share intentions.  Indeed, by a firm agreement they agree – they undertake – to be bound.  It’s a bargain.  Having achieved agreement, each person has a logical right to act as if the other(s) will comply with their common bargain.  One man’s right is another man’s obligation.  We may write the identity as pacta ≡ servanda. 

National laws recognise a law of contract, but the legal system does not, as a rule, become involved with the execution of contracts.  We don’t speak of the weaving and delivery of cloth, or the building of homes, or the bottling of fine wines as legal activities.  Of course they are not illegal, they are legitimate but they don’t require the intervention of the legal system.

In the ordinary way, business doesn’t require the intervention of the legal system, unless someone fails to comply with his bargains.  So it is with arbitration.  The legal system need not, and does not, intervene unless someone, be it a disputant or an arbitrator, fails to do as he or she agreed.

I say, “. . . fails to do as he or she agreed.”  One can say that the words work at two levels: first, the agreement to provide the wine, the house, or the cloth, and secondly the agreement within the agreement, the agreement to arbitrate.  Important though the first may be, at the very root and heart of the relationship between the disputants – they’re often called the Parties – we deal here with the arbitration agreement.  Whoever invokes the Arbitration Agreement is by custom the Claimant; the Party who responds and may make a Counterclaim is called the Respondent.

By the way, the idea behind arbitration is that, because it results from an agreement between two private persons (who may be individuals – real people – or legal persons such as companies or even Governments – but therein lies another tale) the classic form of Doe vs. Haddock is inappropriate.  We speak of an arbitration between Richard Roe and John Doe.  The distinction is subtle but of crucial importance.  It’s the distinction between trial by combat and seeking the help of a friend[1].  Lawyers, in particular, tend to lose sight of this distinction.  When taking off one’s wig, one should also unbuckle one’s sword.

Be that as it may, I argue that Arbitration Law is concerned with two matters and two matters only.  Those matters, for me are (1) process and (2) awards.

(1) Process.

By process, I have in mind to include all those things that must be done if disputants are to obtain the fruits of their arbitration agreement.  Parties to an arbitration are entitled to be treated courteously and any arbitrator should: (a) act fairly and impartially as between the parties, giving each party a reasonable opportunity of putting his case and dealing with that of his opponent, and should: (b) adopt procedures suitable to the circumstances of the particular case, avoiding unnecessary delay or expense, so as to provide a fair means for the resolution of the matters falling to be determined.

The previous paragraph is lifted, almost verbatim, from the English Arbitration Act 1996 but, law or no law, I argue that it stands to reason.  There is much flesh to add to the bones, perhaps too much – the English Act has four parts, four schedules, and one hundred and ten sections; Gary Born – a distinguished legal writer – wrote the Second Edition of International Commercial Arbitration, an authoritative 3600 page treatise, in three volumes,“Lector si monumentum requiris circumspice.[2]; others have produced comparable tomes.  A lot for a simple agreement.

The point remains: arbitration procedure progresses privately between arbitrators and disputants who may, but need not, be represented by Counsel, unless and until one or both Parties decides to invoke the intervention of the Court.

Many arbitrations take place and are completed without attracting the attention of any court.  Some of those – especially in specialised markets and sometimes in construction – are conducted by arbitrators who have no legal training.  Indeed, one cocoa market known to me has a rule that only current traders in that market may be arbitrators at first instance.

(2) Awards

The decision of the arbitrator or arbitrators – the end of the process: end in both senses, the object and the termination – is the Award.  In days gone by and in some territorial jurisdictions, it was a bare decision – who should pay, how much, and to whom.  In contrast with the bare decision is the Speaking or Reasoned Award[3].  There is a great deal to be said about the Award but this essay is not the place or time.

The definitive word on Awards owes everything to logic and little, if anything, to law.  It was by Lord Donaldson in 1981:

“No particular form of award is required.  All that is necessary is that the arbitrators should set out what, on their view of the evidence, did or did not happen and should explain succinctly why, in the light of what happened, they have reached their decision and what that decision is. This is all that is meant by a “reasoned award.[4]””

When I say that Lord Donaldson’s dictum owed everything to logic, I argue that he describes exactly what participants logically can expect from the nature of the arrangements they made.  What concerns me for the purpose of the present essay is what the product is and what the owners can do with it.

I use the word “owners” advisedly.  An award isn’t a public document, although the word “publish” sometimes is used to speak of what the English Act calls “notification of award”.  The arbitrator or arbitrators are the authors but they have undertaken the task at the behest of the Parties and they have been paid, often handsomely.  There may be a question about copyright but, no doubt, others have discussed that.  The only persons who have an interest in the award are the Parties who commissioned it and particularly the Party who succeeded (I find the words “won” and “lost” inappropriate in arbitration and other ADR).

I began by saying that the award had no coercive force.  It can be used to settle with the other Party.  That is all, of itself, an award can do.  A snark’s life may be threatened with a railway-share; an arbitral award does not have even that power.

An award may be taken anywhere.  In theory, any Court in any country may be asked to adopt what it says, whether as a matter of contract or by the operation of that country’s arbitration law, or by the operation of the New York Convention 1958.  In practice there are two courses open if a Party fails to accept and act on an award: one to seek enforcement in the Court of the juridical seat of the arbitration; the other enforcement in a Court of a state that adheres to the Convention.

To jurists it may seem odd that the Court of the seat may purport to annul an Award but only within its own shores.  Another country may see things differently; while Courts respect one another, they do not have to agree[5].  Indeed, although Article V of the New York Convention 1958 creates a presumption that the foreign Court will respect a decision of the Court of the seat of an Award, the word “may” implies that there is no compulsion to do so.

Arguably, it is not the Award that has coercive force any more than any other private arrangement has coercive force.  It is the State and its Courts that have authority to coerce a Party who has not complied with his bargain.

The arbitration legislation of most, if not all national jurisdictions, provides for the more or less direct enforcement of awards made according to certain minimum standards, set out in the two paragraphs under Process and Awards above.  If an award has been made to those standards, there is no need for a Court to rehear the facts and the arguments; the award is enforced by the Court as if it is its own.

It is only in a Court that the Award acquires legal effect.  Unless an application is made to a Court, whether for enforcement or rejection, it remains no more than a record of the decision of a group of private persons, capable of affecting only the Parties who convened the group and appointed them Arbitrators.

Law binds everyone within its jurisdiction with coercive power.  Legal commentators say it has the Jus – the power to declare the law for everyone – and the Imperium – the power of coercion.  Arbitrators have no such powers.  They can decide for their Parties – but no-one else; they have no power of coercion.  Because their decision is not made with the authority of a Judge it has no validity outside the bond of the arbitration agreement.  It cannot therefore become a legal precedent.

Of course great jurists, sometimes themselves great and distinguished Judges, may be appointed arbitrators but they do not take their caps, wigs, or robes into the room.  When acting as arbitrators they set their offices aside.  Arbitrators are merely persons.  They may be great scholars, they may be persons selected at random.  In most jurisdictions the only qualifications are selection by the Parties – or by their agreed method – and compliance with any agreed Rules and the appropriate law of arbitration.  What Court would find itself bound by the word of the present writer, a layman by any criterion?

In conclusion, I submit that there can be no single system of arbitration, transnational or otherwise.  There is autonomy, but a necessarily capricious autonomy of individual arbitrations, each complete unto itself and impinging upon the legal system only in default.

I dare to go further than Lord Mance: Arbitration is not, and cannot be, a law unto itself.  Arbitration’s efficacy as a dispute-resolution mechanism depends upon the individual autonomy of each case – determined upon its own facts and circumstances.  In that sense, Arbitration Law – the cases heard in Courts – may have a jurisprudence but Arbitration per se does not.  The proposition that Awards should be consistent necessarily must be qualified by the facts of each instance; for outcomes to be constrained by external factors not known to the particular disputants would, I argue, be contrary to Natural Justice.

In that opening paragraph, Lord Mance wrote:

“Arbitration already faces problems in maintaining coherence in its jurisprudence and confidence in its efficacy as a dispute-resolution mechanism, particularly given that no general means exist to ensure that awards are consistent.”

I argue that it is the very fact, that arbitration is not constrained to a jurisprudence, and that no general means exist to ensure that awards are consistent, that leads to its inherent flexibility.  It is that flexibility which should give confidence in its efficacy as a dispute-resolution mechanism but not as a source of law.

Arbitration has no need of a declaration of unilateral independence.  Indeed, as there is no entity whose independence could be declared, the concept is a logical impossibility.  Each individual arbitration, however, is independent and autonomous unless and until one party fails to comply with his bargains and the other seeks recourse to the coercive authority of a State.

Arbitration, as a private process, has many advantages, but it cannot usurp the role of the Court.  It cannot proceed ex parte without notice[6].  Its interim measures are ineffective against third parties.  Both are examples of areas in which arbitral institutions have sought to flex muscles that do not exist.  Arbitration appears at first sight to be an alternative to the court but, where coercion is necessary, there is no alternative.

Geoffrey Beresford Hartwell
http://www.arbitrator-engineer-gbh.co.uk/

[1]              Up to a point, Lord Copper.  An arbitrator may be a friend to both parties, but in serious cases arbitrators tend to be selected for their neutrality, among other virtues.

[2]              Inscription in St Paul’s cathedral, London: “Subtus conditur huius ecclesiæ et vrbis conditor christophorus wren, qui vixit annos ultra nonaginta, non sibi sed bono publico. Lector si monumentum requiris circumspice Obijt XXV Feb: An°: MDCCXXIII Æt: XCI.             ”

which translates from Latin as:      Here in its foundations lies the architect of this church and city, Christopher Wren, who lived beyond ninety years, not for his own profit but for the public good. Reader, if you seek his monument – look around you. Died 25 Feb. 1723, age 91.

[3]              See my article, The Reasoned Award in International Arbitration http://tinyurl.com/p33wbj9

[4]              See the full judgment at http://tinyurl.com/lrxxmu6.

[5]              See for example a) Omnium de Traitement et de Valorisation S.A. V. Hilmarton Ltd. [1999] 2 Lloyd’s Rep. 222 Queen’s Bench Division(Commercial Court) before Mr. Justice Timothy Walker; b) France / 23 March 1994 / Cour de cassation / Société Hilmarton Ltd v Société Omnium de traitement et de valorisation (OTV) / 92-15.137

[6]              There are agencies among the many ‘institutions’ that have proliferated – especially since the Second World War – in international and domestic arbitration that have sought to find ways to act surreptitiously without the knowledge of one of the parties to an agreement.  That debate is for another day!

The main part of the title is taken from an article by Nicholas Peacock, Donny Surtani, and Pritika Advani, of Herbert Smith Freehills  – here is a link to the article.   The argument, I suggest, is not restricted to India.  Every time the crime of fraud arises in arbitration, whether in Claim or Counterclaim or even in incidental evidence or argument, an arbitrator has to look into his or her appointment and, indeed, conscience, to know what to do.

I have argued long and often that arbitration is no process at law.  Arbitration has no role to play in criminal law.  An arbitrator’s decision can neither convict a criminal nor can it give effect to truly criminal acts and intentions.

Like any other proposition, any assertion of fraud relevant to the outcome of an arbitration must be tested and assessed according to the balance of probabilities, that is to say that, to succeed,  they must be found more probable than not.  Whether or not there is fraud alleged is a private matter within the arbitration and it is submitted that the mere fact of fraud and its consequences in the arbitration can be decided by the arbitrator or arbitrators,

There is no direct link between the private process of arbitration and the public processes of the criminal courts.  A decision of fraud in an arbitration is private and remains so.  However, an arbitrator who is satisfied, from the evidence, that a fraud may have been committed, may have a duty – either to his own country or another country affected by the offence – to report the suspicion.  Confidentiality, if it exists at all, exists from the nature of the arbitration agreement itself.  It cannot override the public duty to report the crime.

It could be sufficient for an arbitrator to rely upon a party or Counsel to report a fraud.  It is possible, but unlikely, that the parties will be complicit in some fraud against a third party.  It is hard to envisage Counsel as being complicit in such a fraud.  However, arbitrators are not concerned with the regulation of party representatives who may be anyone of a party’s choice.  The representatives may themselves be selected to be privy to the fraud.

In my opinion, an arbitrator should be very reluctant to report a fraud unless he or she is certain beyond a peradventure that a fraud has taken place and the perpetrator can be identified.  If there is no more than a suspicion, however strong, the correct action should be to withdraw from the reference.  It may be necessary to account to the parties with an explanation, but that may have to be very carefully worded,

Having said that, the arbitrator may be compelled to report his suspicions, should he or she consider that the is a risk to the person.

If the arbitrator has a duty to report a fraud, is that duty owed to the seat of the arbitration, to the domicile of one or other or both parties or to the place of performance of the contract (as when constructing a building)?

Note of Proposed Findings

No, it isn’t.  This is not a Note of Proposed Findings; it’s a note about the Note.  I’d better call it the NPF hereinafter.  It’s probably not best to have finished the last sentence with a preposition – if preposition it is – be it never so long but, as I often quote, “the moving finger writes and having writ moves on; nor all your piety nor wit can lure it back to cancel half a line, nor all your tears wash out one word of it” [the Rubaiyat of Omar Khayyam].

That’s especially true of Arbitral Awards, which is why I have a disinterested friend – in confidence of course – read mine before they are published to the Parties.  But perhaps it is literally true when the keyboard is replaced with a touchscreen.  Did Omar Khayyam anticipate the age of the tablet and smartphone when he wrote the verse?

Be that as it may, the NPF originated in the practice of Scottish arbitration.  When Parties desired to have an arbitration award that could be reviewed by the Court, they could ask for the award to be made in the form of a case stated for consideration.  It was helpful if the arbitrator or arbitrators produced an NPF so that Parties could ask for the award to take that due form.

A useful feature of the NPF was that it enabled a party to correct any mistaken impression of the evidence or argument.  Not that if drafted with care it would allow aspects of the matter to be reheard.

There seems to be no legal reason why an arbitrator should use the NPF more widely in most jurisdictions.  I suggest that it could be especially useful where one Party, perhaps tactically or through lack of understanding, has absented itself from the process and the Arbitrator has been appointed by default.  After all the details the recalcitrant or ill-advised Party will have been given, the NPF will become a final warning that absence will have consequences.

Some may object that an NPF would be giving a recalcitrant Party a second bite of the cherry and that would be unfair to the participating Party.  Arbitration isn’t a Court, however, where he who plays the better game wins.  Of course, in Court a Default Judgment may be overturned if an Appeal is successful.  Arbitration, however, is the product of an agreement in which one implication is that both parties intend the arbitrator or arbitrators to find the right solution to their mutual problem.  The Award, once written, is final, it might as well be carved in stone.  In the premises, what can be wrong with giving that absentee one last chance to say his piece before the die is cast – and the metaphor well and truly mixed?

It’s not a situation with which I have had to deal, but I would expect to consider a reasoned application before permitting the absentee to present evidence and argument.  That evidence and argument would be confined to that needed to deal with the NPF.  The compliant Party would be entitled to a rejoinder before the Award was made.

There is a possibility that a late Counterclaim might arise.  I don’t think that should delay the Award, which could be a Partial Award, final as to the matters with which it deals.  Directions for disposing of the Counterclaim could be added to the Partial Award.  I think that a Dispositive section in the Partial Award could well be so separated that it was clear what had been decided finally and what remains to be settled.

In an International context, I think it might be as well to give directions for settling the late counterclaimed matters in a separate document.  The Partial Award would then be final for the purposes of the New York Convention 1958 [http://www.hartwell.pwp.blueyonder.co.uk/ nyc_text.htm] of which Article V.1(e) provides, inter alia, that:

  1. Recognition and enforcement of the award may be refused, at the request of the party against whom it is invoked, only if that party furnishes to the competent authority where the recognition and enforcement is sought, proof that:

. . .

(e) The award has not yet become binding on the parties, or has been set aside or suspended by a competent authority of the country in which, or under the law of which, that award was made.

The first subclause of this provision is the basis for the assertion that an interim decision will not be legally enforced but it should be noted a) that the burden of proof is upon the party against whom recognition and enforcement of an Award is invoked and b) that even then the word “may” suggests that the Court has to be satisfied that its discretion should be exercised.  The curious wording of the Article is worthy of attention.  Scholars argue the distinction between recognition and enforcement, but I think it noteworthy that the Convention uses the singular phrases “it is” and “is sought”, denoting the two as a singular practical concept.

My argument is that, in the context of arbitration – or, for that matter, adjudication or expertise – it remains perfectly proper for the decision-maker to say to the Parties, and particularly any absentee, “On the basis of what you have told me I am minded to find thus and so”, making it clear that his or her mind remains open and flexible.  The NPF is a formal representation of that principle.