Giving Contractual Notice by E-mail

July 9, 2008

Contracts often provide for notices to be sent, such as to trigger an event in the contract. There has always been some risk surrounding the giving and receiving of notices. For example: what if a notice is sent to the wrong address? What if a notice is lost en route? What if the intended recipient does not read a notice?

Faced with such questions, the courts have ruled over the years to give practical commercial effect to advancements in communications such as the postal system, telexes, telephones and faxes, and deal with new technological challenges as they emerge. A relatively recent development is e-mail. Like previous technological advancements, e-mail brings a new set of risks and challenges for commercial dealings.


Freedom of contract

The reported cases on contractual notice are based on principles of contractual freedom, construction and commercial pragmatism. While the courts have imposed specific rules regarding communication of information in some areas, such as in relation to contract formation, few constraints are imposed on parties to contract as they see fit regarding notices. As Lord Hoffman illustratively put it Mannai Investment Co. Ltd v Eagle Star Assurance [1997] AC 749:

‘If [a] clause had said that the notice had to be on blue paper, it would have been no good serving a notice on pink paper, however clear it might have been …’.

The UNCITRAL Model Law and the EU Directive on Electronic Commerce[1] similarly respect freedom of contract with key provisions applying only where nothing is otherwise agreed.

The question of which party should bear the risk of a failure of notice will always turn on the facts. If the risk of giving notice may be considered comparable to the risk of communicating contract acceptance then, as Lord Wilberforce said in Brinkibon Ltd v Stahag Stahl und Stahlwarenhandelsgesellschaft mbH [1983] 2 AC 34:

‘No universal rule can cover all such cases: they must be resolved by reference to the intentions of the parties, by sound business practice and in some cases by a judgment where the risks should lie’.

However, it is important to distinguish the act of giving notice under a contract (the subject of this article) from contract formation. The starting point for rules relating to the communication of acceptance of an offer is that a contract has not yet been formed. On the other hand, the act of giving a contractual notice is, in the absence of any external factors, governed by the contract itself. There is no suggestion, for example, that the reasoning behind the postal acceptance rule is applicable to matters beyond contract acceptance. While contract formation cases therefore cannot be directly applied to contractual notice situations, they are useful references of the approaches taken by the courts.

Right to give notice via e-mail

In the absence of any statutory imposition, whether a notice may be validly delivered via e-mail will depend on the proper construction of the contract.

Where a contract expressly prescribes the permitted methods of notice, the courts will give effect to those requirements (see Mannai Investment Co Ltd v Eagle Star Assurance [1997] AC 749). If a contract expressly requires that a notice be delivered by post to a specific address, it is unlikely that a notice delivered by e-mail would be valid. This would be the case even though it may have actually come to the attention of the intended recipient and therefore achieved the functional outcome of communicating information.[2]

What, then, is the situation if a contract is silent as to the permissibility of delivery of a notice by e-mail but simply provides that the notice be in ‘writing’?

In New Zealand, the question of whether writing includes electronic information is governed by s 18 of the Electronic Transactions Act 2002, implementing Article 6 of the UNCITRAL Model Law. This provides that a requirement for writing may be met by information in electronic form. Importantly, the New Zealand legislation introduced a further provision relating to the giving of information which requires the recipient to consent to receiving the information in electronic form.[3] Therefore, in the case of contractual notice, it is still a question of the intention of the parties.

The evolving ‘customary usage’ of e-mail in commerce will be relevant in interpreting the objective intentions of the parties. With business communications now commonly sent by e-mail, a requirement for ‘writing’ in many cases may be presumed to include e-mail unless stated otherwise. Certainly, in a modern commercial context, where very often the majority of information is in electronic form, it would be difficult to argue that ‘writing’ can never include electronic information.[4]

With such widespread use of electronic information in commercial settings, to the displacement of handwritten and printed material, it is likely that, unless the context requires otherwise, a modern reference to ‘writing’ in a commercial contract will include electronic information such as e-mails.

Similar questions arise in relation to where notice must be delivered. If a contract requires that a notice be ‘delivered’ to a physical address, such as a company’s offices, will an e-mail sent to a person working in that office be valid? Such a requirement could be expressed as a physical address, e.g. ‘123 Vine St’, or by reference to a physical address, e.g. ‘the offices of Vine & Company’.

In New Zealand, this is governed by s 13 of the Electronic Transactions Act 2002, implementing Article 15 subsection 4 of the UNCITRAL Model Law. The section provides that an electronic message is deemed to be received ‘where the addressee has its place of business’, or ‘the place of business … which has the closest relationship to the underlying transaction’ unless otherwise agreed. While these provisions are brief and raise further questions themselves, the likely position is that a notice sent via e-mail to a recipient at a business-related e-mail address will be deemed received at their place of business.

In the absence of such provisions, as in the UK, if the requirement is merely for ‘delivery’ of information and not a physical thing, it can be argued that, once an e-mail has been delivered to a recipient at an e-mail address used by them for business communications and with a sufficient relationship to the physical address for notice (perhaps as evidenced on a company website, directory listing or business card), then the information is deemed delivered to the physical address. This is consistent with the approach implemented in the UNCITRAL Model Law. It can also be argued that the situation is analogous to sending a notice, required to be delivered to a company at a physical address, via letter to that company’s post office box. In such a case the letter is not delivered to the physical address but to a post office box operated by a third party. If the post office box is one known to be used by the intended recipient, it may be deemed delivered to the recipient’s physical address.

Ultimately, whether it is permissible to send a notice by e-mail depends on the circumstances of the contract. In the absence of any indication to the contrary in either the contract or legislation, there is a reasonable likelihood in modern contracts that a right to give notice by e-mail will be implied.

Deemed receipt

If e-mail is a valid method for delivering a notice under a given contract, a key issue may arise as to when receipt of the notice occurs, if at all. The issue of receipt will have an important bearing on where the risk lies in the event of the notice being lost.

In New Zealand, having implemented the UNCITRAL Model Law, the position is relatively clear. Article 15 states that, unless otherwise agreed, where an information system has been ‘designated’ for the purpose of receiving electronic communications, then receipt of a communication occurs ‘at the time when the data message enters the designated information system’.[5]

If an e-mail address is stated to be used for communications, the system that receives and stores e-mail sent to that address (the ‘information system’) would be the ‘designated’ system.[6] This would be the case whether a company hosted its own e-mail system or used an ISP’s e-mail system. It may also be that, if an e-mail address is held out on a document such as a company website, advertisement or business card, the e-mail system used in relation to that address may be ‘designated’.[7]

On this basis, as soon as a notice sent by e-mail is accepted by the intended recipient’s e-mail system, the notice will be deemed received, whether or not it has been retrieved by any user (ie when a user checks e-mail and downloads any new messages). The negation of any requirement for retrieval is emphasised by the subsequent subsection in Article 15 which says that, if a communication is sent to an information system other than that designated, receipt occurs ‘at the time when the data message is retrieved by the addressee’.[8]

Where the conditions are not met, and in the UK which has not implemented the UNCITRAL Model Law, the position is nevertheless likely to be the same; specifically, that receipt of an electronic communication occurs once the communication has been accepted by the intended recipient’s system. It has been held that a notice need not be read in order to be deemed delivered.[9] As Lord Fraser of Tullybelton said in Brinkibon in relation to telex communications:

‘once the message has been received on the offeror’s telex machine, it is not unreasonable to treat it as delivered to the principal offeror, because it is his responsibility to arrange for prompt handling of messages within his own office’.

Similarly, the responsibility to take any other steps as necessary, such as forwarding an e-mail to another person, lies with the intended recipient.

Furthermore, if retrieval of an e-mail were required, it would be conceivable for the recipient to be aware of the notice having been received but be able to refuse to retrieve it. The courts have ruled out such a possibility in relation to postal communications:

The usefulness of a presumption of this kind [that letters correctly addressed and posted arrive at their destination and are read] would be destroyed if the addressee could nevertheless be heard to say: “Although I received the postal packet quite safely, I did not read the content” or “I did not examine the postal packet to see that I had extracted all that it contained”.[10]

Such reasoning can be neatly applied to e-mails. Therefore, the logical point for receipt prior to actual retrieval is when an e-mail is received by the intended recipient’s system. At this point, the message has entered the ostensible control of the recipient and there is nothing more the sender need do. Giving effect to this consideration is consistent with the commercially pragmatic approach taken by the courts. It is also consistent with comments of the England and Wales Law Commission in its 2001 report ‘Electronic Commerce: Formal Requirements in Commercial Transactions’. The commission reached an ‘initial view’ that ‘the transmission process is complete when the communication reaches the recipient’s Internet Service Provider’, although they declined to make any further recommendation on the issue of deemed receipt.[11]

The EU Directive on Electronic Commerce takes the same approach in relation to placing of orders by electronic communication, which are ‘deemed to be received when the parties to whom they are addressed are able to access them’.[12]

Therefore, there is a reasonable likelihood that delivery of a notice sent via e-mail is completed once it is accepted by the intended recipient’s e-mail system. Accordingly the risk is likely to lie with the recipient from that point onwards in the event of the e-mail being subsequently lost.

What can go wrong?

Once an e-mail has been successfully received by a recipient’s e-mail system, there are numerous reasons why it can still be ‘lost’, and therefore possibly never read by the intended recipient. Loss may occur ‘silently’ with no notification being sent and little if any evidence of what happened. This is significant because, as argued above, once an e-mail has been accepted by the recipient’s system, it will be considered delivered whether or not it is ever read.

Some of the possibilities by which e-mail could be received but never read are as follows.

Anti-spam software

It is possible that a notice sent by e-mail could be automatically deemed as spam and discarded.

Many companies and ISPs now use software to filter all incoming mail for spam. There is a wide variety of anti-spam software employing many different techniques to detect and block spam. Techniques include using the presence of key-words to instantly flag an e-mail as spam or using mathematical algorithms to ‘intelligently’ identify spam. Other systems analyse not the content of the e-mail but the origin, with e-mail sent from certain addresses being blocked. Most systems allow a wide range of user configuration as to how spam is detected and the actions the system will take following detection.

A common factor in systems which detect spam based on content is that an e-mail must be ‘received’ by the system in order to be analysed. If the e-mail is determined to be spam, it may be redirected, flagged, ‘bounced’ back to the sender or any other action taken, including being discarded without notice, as configured by the software. However, by this time the e-mail will have already been deemed received.

System failure

After an e-mail has been received by the recipient’s system, a malfunction such as a hardware fault or software bug could result in the e-mail being lost before it is read. There are numerous points of failure depending on what is taken to be the recipient’s ‘system’ and the various hardware and software components in question.

Even in a stable and trusted system, problems may occur. Software bugs may be introduced when updating the existing software. Best practice requires that software be kept up to date in order to fix security flaws and other important issues as they are found and fixed over time. An updated system may ‘break’ the existing configuration causing e-mails to be lost. It is even possible that an external event could result in a system failing. For example, a reassignment of IP addresses by an ISP may cause an e-mail system located on another computer to malfunction.

Human error

Human error, such as a misconfiguration or inadvertent deletion, could result in an e-mail being silently ‘dropped’ or otherwise lost after being received by a system but before being read. Intentional damage by hackers poses a risk, as does malware such as viruses. While there are comparable risks with postal mail and other forms of communication, the intangible nature of e-mail means that such events can happen instantly and without trace.

Unchecked mailboxes

E-mail systems use the concept of virtual mailboxes into which messages are organised. If an e-mail is delivered to a mailbox which, for any reason, is never checked by a user, then the e-mail will likely never be read.


E-mail might be automatically redirected based on certain criteria. This may be as the result of a user’s own actions or a system-based rule. One such scenario is that a user might set up a rule to automatically forward particular e-mails to another address. If the user inadvertently specifies the wrong forwarding address, then e-mails may be lost without being read.

Comparable risk

The scenarios listed above could result in a notice sent by e-mail being deemed to be received but, through no fault of the sender, never being read by the intended recipient. These are risks associated with electronic communication, and they are comparable (though not equivalent) with the risks of postal communication. For example, a letter may be mishandled within an office or physically destroyed by fire.

There is also comparable risk of loss for both e-mails and letters preceding receipt. In the case of e-mail, a message may have to pass through many third-party systems before entering the recipient’s system with a possibility of loss at each step. In the case of the postal system, a letter may be lost or stolen in the post. Royal Mail reported that in the 2007/2008 period 0.34% of mail items were not ‘delivered correctly’, and in the previous year received 651,582 complaints of loss. While this is statistically a very low level of failure, out of the billions of items posted every year thousands will be lost.

This risk has been commercially accepted, with postal communications and more recently e-mail being routinely used in everyday business. In cases where a higher degree of certainty is required, more reliable methods such as registered post may be used. Some firms are now offering ‘registered e-mail’ services which claim to offer proof of delivery.[13]

The courts are unlikely to impose different rules on notices sent by e-mail than on notices sent by post. Similarly, electronic commerce laws such as those based on the UNCITRAL Model Law aim to provide consistency and compatibility with existing forms of communication. To that end, the principle of ‘functional equivalence’ has been used to describe the approach of extending rules to electronic communications.[14]

Where should the risk lie?

As discussed, there are many scenarios by which an e-mail could be ‘lost’ after having been received by the recipient’s system and therefore never read. Should the risk lie with the sender or the intended recipient in such cases? More specifically, should the intended recipient be able to say that because he did not read the e-mail and was not aware of its being sent, he is not liable for any resulting loss?

It is established that notices do not need to be read in order to be effective and there is a presumption of ‘prompt handling’ of messages within an office[15] – a recipient cannot use ‘administrative’ errors within their business as a defence. It is consistent to say that neither can a recipient use ‘technical’ errors within systems as a defence.

This proposition is supported in Entores Ltd v Miles Far East Corporation [1995] 2 QB 327 (affirmed in Brinkibon), where Denning LJ (obiter) gave the example of a telex machine that had run out of ink and failed to print a message otherwise successfully transmitted. In such a scenario, where there was fault on the part of the receiver (eg failure to replenish ink in a telex machine) but no fault on the part of the sender, the recipient would be estopped from denying receipt of the message. Denning LJ went on to say that if there was no fault on the part of the recipient (the offeror in the case) then the recipient would not be estopped from denying receipt. Where an e-mail has been lost from the recipient’s system, it is probable, at least in the scenarios noted earlier, that the loss is attributable to some fault on the part of the recipient or their agent. For example, a misconfiguration, failure to keep backups, etc, whether by the recipient or their agent, can be seen as modern equivalents to a failure to keep an adequate supply of ink in a telex machine.

Furthermore, to allow a distinction between ‘administrative’ and ‘technical’ errors gives rise to problems of classification. For example, if an e-mail is lost due to a misconfiguration of the e-mail system, would that be a ‘technical’ error or an ‘administrative’ error? It is unlikely that the courts would be drawn to impose different rules based on such fine distinctions.

It is unlikely to be a defence for an intended recipient to claim that an e-mail was not read due to a ‘technical’ reason beyond the control or knowledge of the sender. The onus is firmly on intended recipients to ensure that e-mails are handled properly by their computer systems and personnel at all times.


There is sufficient basis to conclude, applying common-law principles, that where a notice may be given by e-mail it will be deemed received when it enters the intended recipient’s e-mail system. Companies proposing to receive notices by e-mail should therefore be aware of the risks this presents. Furthermore, in the absence of an indication to the contrary, a modern contract may be taken to contain implied consent to receive notices by e-mail.

In order to reduce the risk posed to recipients, contracts should expressly define the valid form of notices, including whether and how notices may be sent by e-mail. If notices are not intended to be sent by e-mail, the contract should explicitly state that notices must be in writing and the address to which they must be posted or otherwise delivered. Beware of contractual definitions that include e-mails as writing.

Consider the following when reviewing a contract where notices may be sent by e-mail:

l        Expressly state the e-mail addresses to which a notice must be sent.

l        Expressly state that notices sent by e-mail are not received until retrieved and read during normal business hours. The client may also consider imposing further conditions to e-mail notices, such as requiring a physical copy to be posted or confirmation of receipt in order to be valid.

l        Check for local legislation which may impose deeming provisions or other rules in relation to electronic communication.

l        Advise the client to check that their e-mail system works to a high degree of reliability for the relevant e-mail addresses, including understanding how any spam filters may block messages based on certain rules.

l        Suggest to the client that, if possible, the relevant e-mail address be set up as a ‘distribution list’ to copy the message to more than one recipient.

l Advise the client to put a process in place to regularly check the e-mail addresses and ensure notices are handled properly once received.


Guy Burgess LLB, BCom is an IT lawyer and software developer based in New Zealand. He has acted on international software development disputes and specialises in advising on IT projects and contracting. He is contactable at


[1]    Directive 2000/31/EC ‘Directive on electronic commerce’, June 2000.

[2]    Following the reasoning in Entores Ltd v Miles Far East Corporation [1955] 2 QB 327 and Mannai it is arguable that in some cases there may be an onus on the recipient to advise the sender of such an error in notice.

[3]    Electronic Transactions Act 2002, s 20.

[4]    For a similar view see the England and Wales Law Commission 2001 report, ‘Electronic Commerce: Formal Requirements in Commercial Transactions’, Part 3, regarding statutory requirements for writing.                   

[5]    Article 15, Section 2(a)(i). Implemented in the Electronic Transactions Act 2002, s 11.

[6]    See Guide to Enactment of the UNCITRAL Model Law on Electronic Commerce, para 102.

[7]    The Guide to Enactment of the UNCITRAL Model Law on Electronic Commerce at para 102 states that the mere indication of an e-mail address on such a document ‘should not be regarded as an express designation’ (emphasis added). However it may be argued to be an implicit designation.

[8]    Article 15, Section 2(a)(ii). Emphasis added. Implemented in the Electronic Transactions Act 2002, s 11(b)(ii), though with different wording.

[9]    Brinkibon; Stidolph v American School in London Educational Trust Ltd (1969) 20 P&CR 802; Anson v Trump [1998] 1 WLR 1404.

[10]  Stidolph v American School in London Educational Trust Ltd (1969) 20 P&CR 802, 805, per Edmond Davies LJ.

[11]  Paragraph 3.56 (1).

[12]  Directive 2000/31/EC ‘Directive on electronic commerce’, June 2000. Emphasis added.

[13]  E-mail systems often allow for automatic responses to be sent, such as a ‘delivery receipt’ indicating an e-mail has been delivered to the recipient’s system, a ‘read receipt’ indicating that a user has read an e-mail, and a ‘non-delivery report’ to inform the sender of a problem delivering an e-mail. However, these cannot be relied on as determinative. There is no guarantee as to how any given e-mail system will implement automatic responses, or even whether they are enabled at all.

[14]  See ‘Guide to Enactment of the UNCITRAL Model Law on Electronic Commerce’, Part I, section E.

[15]  Brinkibon.