A Question of Causation: Knowing
Assistance and the "Duty to Inquire"

Simon Baughen*

After the decisions in AGIP (Africa) Ltd. v. Jackson and others I (AGIP) and Lipkin
Gorman v. Karpnale Ltd. and another2 (Lipkin Gorman), dishonest assistance in a
dishonest breach of a fiduciary relationship seems to be what is required to establish
liability for knowing assistance.3 In assessing the state of a defendant's knowledge
reference can be made to the five "Baden" categories of knowledge. 4 The first three
categories indicate dishonesty:
(1) the defendant had actual knowledge of the fraud
(2) the defendant wilfully shut its eyes to the obvious,
(3) the defendant wilfully and recklessly failed to make such inquiries as an honest

and reasonable man would make.
The last two categories indicate only negligence:

(4) the defendant knew of circumstances which would indicate the facts to an honest
and reasonable man,

(5) the defendant knew of circumstances which would put an honest and reasonable
man on inquiry.

What is less clear, however is when the relevant "dishonesty" will be found on
the facts. This article seeks to examine two areas of uncertainty arising out of the
question of causation. The first concerns a defendant who suspects, but does not actually
know, that it is giving assistance to a dishonest scheme but dares not seek clarification
for fear that its suspicions might be confirmed. Can such a defendant argue that it
is not liable for knowing assistance on the grounds that any questions it had asked

* Lecturer in Law, The University of Bristol. The author wishes to thank Mrs. Helen Norman of the
University of Bristol who kindly commented on a previous draft of this article. The views expressed in
the article are those of the author alone.
I. [1991] 3 W.L.R. 116.
2. [1991] 3 W.L.R. 10 (HL). Of particular relevance to the issues addressed in this article is the judgment
of the Court of Appeal, [1989] I W.L.R. 1340.
3. The Court of Appeal's decision, given by Fox L.J., in AGIP must qualify this assertion to some extent.
At p. 131 G he refers to the relevant test of knowledge as that set out in Selangor United Rubber Estates
v. Cradock (No 3) [1968] I W.L.R. 1555, namely, "circumstances which would indicate to an honest
and reasonable man that such a [fraudulent] design was being committed or would put him on inquiry
as to whether it was being committed". However, his subsequent analysis of the facts makes it clear
that dishonesty is the relevant factor. At p.l13 G he states, ''The question is whether [they] acted honestly".
4. See Baden. Delvaux and Lecuit v. Societe General pour Favoriser Ie Developpement de Commerce
et de [,Industrie en France S.A .. [1983] B.C.L.C. 325.

27



THE DENNING LAW JOURNAL

would have been answered in such a way as to allay its suspicions? The second concerns
the extent to which the plaintiff's own negligence can bar or limit its right to sue
a defendant for knowingly assisting the fraud.

1. The "duty to inquire" and causation.
(a) A difference of approach in AGIP and Eagle Trust?

This issue appears to be subject to two differing views at first instance (the issue
was not considered in Fox L.J.'s judgment in AGIP). In AGIP Millett J. strongly
rejected the idea that a dishonest defendant could raise the defence that any further
inquiries it might have made would have failed to flush out the fraudster. The
defendant's argument was based on the judgment of Peter Gibson J. in Baden5 which
discussed the position of a party who is put on inquiry but fails to inquire and concluded
that the plaintiff must prove that the inquiry would have disclosed the truth; the burden
lying on it to prove a causal connection between the failure to make inquiry and the loss.

Millett J. vigorously rejected the idea that this argument had any relevance to cases
of dishonesty as opposed to constructive notice. In his view the argument

"derive [d] from a misunderstanding of the basis of the constructive trustee's
liability. He is not liable for failing to make inquiry, but for the misapplication
of the plaintiff's property. He is under no duty to make inquiry, his only duty
is to act dishonestly. If he makes inquiry, he does so for his own protection.
If he does not make inquiry, the loss is not caused by his failure to do so but
by his participation in the misapplication of the plaintiff's funds. He is liable
only if he acted with knowledge; and this must be judged in the light of all the
circumstances known to him and any explanation actually given to him. But
it is not, in my view, to be judged considering the hypothetical explanations
which might have been given to him ifhe had sought them. If it were otherwise,
his liability would depend on whether the fraudster would have been sufficiently
inventive to be able to supply a plausible explanation if asked for one. In the
present case it would depend on whether the defendants should be assumed to
have directed their inquiries, which ex hypothesi they did not make, to Mr. Zdiri
[the fraudster] or to his superiors. Such considerations are or ought to be
irrelevant. "6

In contrast, Vinelott J. in Eagle Trust pic v. SBC Securities Ltd. appeared to suggest
that such a line of enquiry is relevant to the issue of dishonesty. 7 Dealing with the
allegation that dishonesty was borne out by a failure to make inquiry, he said:

"Savory Milln may have felt some anxiety, and even entertained some suspicion,
as to how Ferriday had managed to arrange for the £13.5m to be paid by Anser

5. Ibid.
6. Supra.n.1 at pp. 295 G-296 C.
7. [1992] 4 All E.R. 488.

28



A QUESTION OF CA USA TION: KNOWING ASSISTANCE AND THE' 'DUTY TO INQUIRE"

on his behalf. But it is to my mind going altogether too far to say that any honest
and reasonable man would either have inferred that, despite the denial to Eagle's
solicitors, he was indemnifying the sub-underwriters or some of them on behalf
of Eagle or that the £13.5m was Eagle's money which had been misappropriated
by Ferriday in a gross breach of his duty as a director and in contravention of
s.151; or that an honest and reasonable man would not have applied the money
in discharge of the liabilities of Earnshaw Haes and Ferriday until he had been
satisifed by inquiry that the moneys were properly paid. It is not easy to see
what further inquiries Savory Milln could have made. If inquiry had been made
of Ferriday and he had been innocent of any wrongdoing, he might well have
resented the inquiry; and if guilty, it is hardly likely that, faced with a second
inquiry, he would have confessed at all. And . . . if he had refused to answer,
what further inquiry could Savory Milln have made? It would hardly have been
open to them to invite Eagle to play detective and investigate their own chief
executive. "8

(b) The facts of AGIP.
At first glance, the above passages suggest a conflict of judicial opinion as to the

relevance of causation and the duty to inquire. However, a closer examination of
the facts of those two cases suggests that this conflict is more apparent than real.
Of particular significance, it is submitted, is the state of the defendant's knowledge
at the time the so-called "duty to inquire" arose.

In A GIP the fraud arose when Mr. Zdiri, a senior officer of the plaintiff, fraudulently
altered the name of the payee on a payment order, after it had been signed, in favour
of Baker Oil. Baker Oil had recently been incorporated and its two directors and
shareholders were a partner of Jackson's and an employee. Thereafter the money
was transferred to Jackson's account and was eventually transferred overseas to various
recipients. At the time Jackson's effected these transfers, Millett J. stressed that they
already had the requisite guilty knowledge, therefore any "duty to inquire" would
strictly be a misnomer. The purpose of any such enquiry would be purely for Jackson's
own protection.

Millett's finding was based upon the following facts. For the two years preceding
the fraud some US$ 10m had come from the plaintiffs to Kinz, via companies set
up through Jackson's, which ran a jewellery business in France. Jackson's instructions
in these transactions came from the recipients and not the plaintiffs.

"They knew of no connection or dealings between the plaintiffs and Kinz or
of any commercial reason for the plaintiffs to make substantial payments to Kinz.
They must have realised that the only function which the payee companies or
Euro-Arabian performed was to act as "cut-outs" in order to conceal the true
destination of the money from the plaintiffs. They must also have realised that

8. Ibid. at p.511 A-D.

29



THE DENNING LAW JOURNAL

the only purpose in having two' ,cut-outs" instead of one was to make impossible
for investigators to make any connection between the plaintiffs and Kinz without
having recourse to Lloyd's Bank's records; and their object in frequently replacing
the payee company by another must have been to reduce the risk of discovery
by the plaintiff."9

Moreover, a letter of 14th August, 1984 from Knapp-Fisher showed that Mr Jackson
was concerned at the possibility that the plaintiffs might obtain disclosure of Lloyds
Bank's records, discover what had happened to the money and try to recover it. Against
this background, Millett J. found that Jackson and Griffin' 'obviously knew that they
were laundering money. They were consciously helping their clients to make
arrangements that they were laundering money . . . They must have realised that
their clients might be involved in a fraud on the plaintiffs."10

Although Jackson and Griffin never gave evidence, the external evidence, such
as the Knapp-Fisher letter, pointed to the fact that they may have believed that the
fraud being committed was a scheme to avoid Tunisian exchange control, possibly
with the connivance of the plaintiffs. However, Millett J. attached no significance
to the fact that Jackson and Griffin may have got the details of the fraud wrong. What
mattered was that they must have known that the banking arrangements set up through
themselves could only have been established with a view to perpetrating a fraudulent
purpose. This would be enough to establish the necessary dishonesty on their part
to make them liable to the plaintiffs in an action for knowing assistance.

Millett J. also went on to mention the fact that the defendants never made any enquiries
or took any steps to satisfy themselves that the arrangements had the plaintiffs'
knowledge and approval. In the light of his subsequent comments on the issue of
causation, however it is clear that he did not regard this factor as being essential
for establishing dishonesty. Effectively, Millett 1. was treating this as a "Baden 2"
case rather than a "Baden 3" one.

Despite Fox L.J.'s unfortunate use of the Selangor test for knowledge, it is clear
that he regarded the crucial issue as being whether or not the defendant had acted
honestly. In this connection he concluded that Millett J. rightly carne to the conclusion
that the defendants must have known they were laundering money and "were
consequently helping their clients to make arrangements to conceal some dispositions
of money which had such a degree of impropriety that neither they nor their clients
could afford to have them disclosed. "11 Like Millett J. he placed particular emphasis
on Knapp-Fisher's letter to Jackson's of 14th August, 1984:

"Mr. Jackson and Mr. Griffin were being given advice on the possibility that
a payment or payments might involve a fraud on AGIP. Having got to that point
it seems to me that persons acting honestly would have pursued the matter with

9. [1989] 3 W.L.R. 1367 per Millett J. at 1390.
10. Ibid.
II. Supra. n.l at p.133 A.

30



A QUESTION OF CA USA TION: KNOWING ASSISTANCE AND THE' 'DUTY TO INQUIRE"

a view to satisfying themselves that there was no fraud. But there is nothing
to show that they did that. They made no inquiries of AGIP at all. They let the
matter continue." 12

Like Millett J., Fox L.J. referred to the defendants' failure to make inquiries, but
unlike Millett J., did not go on to examine how critical that failure was in establishing
dishonesty in the defendant.

(c) The facts of Eagle Trust
In October 1987 the plaintiff company made a take-over bid for a group of companies.

The defendant company agreed to underwrite the cash alternative and the rights issue
which formed part of the offer. The defendant sub-underwrote its liability, using
a list of underwriters introduced by plaintiff's chief executive, Mr. Ferriday. The
list included Ferriday himself as underwriting 25.5 million shares. The plaintifflater
claimed that the defendant ought to account for £13.5 million paid under the sub-
underwriting arrangements. It alleged that Mr. Ferriday had fraudulently arranged
for its money to satisfy obligations, including his own arising out of the sub-
underwriting, and that the defendant ought to have known of this or have been put
on enqUIry.

In this connection two particular facts were relied on. First, as Mr. Ferriday's
wealth was largely tied up in Eagle, after "Black Monday" it ought to have been
clear that he would have lacked the funds necessary to satisfy his sub-underwriting
obligations. Secondly, Mr. Ferriday had arranged the sub-underwriting list and had
altered it at a late stage. The allegation was made that, in the case of Martin Boston
& Co.lMartin Boston, the list was changed after they had denied they were liable
and their obligation had been taken over by Ferriday. However, when asked by Eagle's
solicitors, Mr. Ferriday had denied acting on behalf of that firm.

In contrast to the facts in AGIP, this factual base on its own would be insufficient
to establish that the defendant must have known that Mr. Ferriday was discharging
his liabilities out of Eagle's money; especially in the light of Mr. Ferriday's earlier
denial to Eagle's solicitors. The "duty to inquire" serves a different function than
on the stronger facts of AGIP. There it served only as a possible means whereby
a defendant might extricate itself from pre-existing guilty knowledge. In Eagle Trust
it serves as a means of establishing whether that guilty knowledge ever existed, a
potential "Baden" 3 case. This, therefore, provides a possible basis for reconciling
the two decisions. Causation will be relevant, but only if the "duty to inquire" is
itself necessary to establish guilty knowledge by the defendant.

On closer examination, however, this distinction is unconvincing. In Eagle Trust,
the defendant was not dishonest and therefore was not liable for knowing assistance.
A key factor in this finding was Ferriday's earlier denial to Eagle's solicitors. The
issue is not affected by the question of what additional information further inquiries

12. Ibid. at p.134 E.

31



THE DENNING LAW JOURNAL

could have revealed. It is very difficult to distinguish "Baden 3" knowledge from
"Baden 5" knowledge and it is submitted that Millett J. was correct to state in AGIP
that issues of causation should only be relevant to questions of constructive notice
which have no place in the fraud based action of knowing assistance. Moreover, to
treat as relevant this issue of causation runs the risk of diluting the deterrent effect
of the action on potential fraudulent assisters.

It is unfortunate that this issue has yet to be clarified at Court of Appeal level.
Fox L.J. ignored it altogether in A GIP and the state of the pleadings in Lipkin Gorman
precluded any serious examination of what facts could constitute dishonesty. What
dicta there are in that case tend to confirm the muddled thinking identified by Millett
J. in AGIP. May L.J., criticising Selangor, stated that "it was wrong to equate the
duty to inquire where there has been fraud and the bank is proved to have known
of it with that where all that is being alleged is that the bank has been negligent". J3
With respect, if the bank is proved to have known of the fraud, that disposes of the
issue and any question of a duty to enquire becomes irrelevant. However, the words
used by May L.J. identify the key problem with using a "duty to inquire" to establish
dishonesty, namely that it tends to confuse dishonesty with mere negligence. This
confusion is illustrated by Fox L.J. ' s use in A GIP of the Selangor test which covers
both negligence and dishonesty, whereas his subsequent analysis of the facts is directed
entirely towards dishonesty alone. It is submitted that the law in this area would be
clearer if the concept of a "duty to inquire" were to be abandoned altogether and
knowledge categorised broadly as either dishonest or honest but negligent.

2. Causation, contributory negligence and "clean hands".

In Lipkin Gorman Cass, a salaried partner in a firm of a solicitors abstracted money
from its client account to fund his gambling habits. The solicitor sought to recover
the money from either the bank or the Playboy Club where the gambling had taken
place. The claim against the bank was framed as one for breach of a duty of care
owed to the solicitors as its clients in both contract and tort and also as a claim for
knowing assistance.

Alliott J. found for the solicitors in both claims, but limited the amount of their
recovery to the losses occurring in the period from the breach of duty by the bank
to 1st October, 1980 by which time Cass's fellow partners should have discovered
the full extent of his embezzlement for themselves. By that time one of Cass 's partners
had confronted him with a false claim for travel expenses and had accepted Cass's
explanation that that was all that was owed. However, Cass stayed on as a signatory
and from that date was able to steal two further sums of £2, 102.05 and £ 120,698.61.
From that date, their contributory negligence was 100 per cent, so barring the
contractual action in which liability was the same as in negligence, independently
of the existence of any contract. The knowing assistance claim was also barred on

13. [1989] 1 W.L.R.1340atp.1355H.

32



A QUESTION OF CAUSA nON: KNOWING ASSISTANCE AND THE' 'DUTY TO INQUIRE"

the grounds that it would be inequitable to give them any relief. May L.J. endorsed
this approach on the grounds that the plaintiff could no longer come to the court with
"clean hands" .14

It is submitted that this approach is quite wrong. There is no equivalent to contributory
negligence in actions for breach of trust. The nearest equivalent is a defence that
the plaintiff consented to the breach of trust. This patently does not cover the plaintiff s
conduct in Lipkin Gorman. Nor does the doctrine of "clean hands" operate against
a negligent plaintiff. In Williams v. Staite misconduct by a licensee towards the
freeholder did not affect its claim to an estoppel licence. IS Some element of
dishonesty in the plantiff is required before it ceases to have "clean hands" as with
the attempt to deceive the court with forged documents in J. Willis & Sons v. Willis,
another licence case. 16

On the facts of Lipkin Gorman it is easy to see how both Alliott J. and the Court
of Appeal came to the conclusion they did. On the evidence of banking practice in
that case, some form of dishonesty would be required to establish a breach of duty
of care in contract and tort. Therefore it is easy to assimilate that liability with knowing
assistance and to conclude that if contributory negligence operates on one action some
equivalent must be found for the other. The actions are not identical, however, but
only appear to be so because the effect of the evidence as to banking practice is that
a bank which honours its mandate can never be negligent unless it is also dishonest.
The extended concept of "clean hands" is quite inappropriate to an action based
on the dishonesty of one who assists a fraudster. For instance, if in AGIP the plaintiff
had been careless in allowing Mr. Zdiri to perpetrate his fraud, neither of the defendants
would have been any less dishonest in assisting him. To curtail the plaintiffs recovery
would be to punish it for carelessness and allow a fortuitous escape to the fraudster
and accomplices.

Conclusion
Two sets of muddled thinking have been going on in the knowing assistance cases

discussed above. The first is to allow the similarities between' 'Baden 3" and "Baden
5" knowledge to blur the boundaries between dishonesty and negligence and allow
questions of causation relevant to the latter issue to intrude onto the former. The
second is the evidential quirk in Lipkin Gorman that a bank honouring its mandate
can never be negligent and honest. This has led both Alliott J. and the Court of Appeal
to assimilate the bank's contractual liability to its customer with its liability for knowing
assistance. However, the basis of these actions is quite different and to seek an
equivalent to contributory negligence in knowing assistance is not required on grounds
of consistency and has the undesirable effect of diluting any deterrent effect the action
for knowing assistance may have.

14. Ibid. at p.1360 A.
15. [1979] Ch. 291.
16. [1986] 2 E.G.L.R. 62.

33