In X v Twitter Inc [2017] NSWSC 1300 the Supreme Court of New South Wales, per Pembroke J, issued a final injunction regarding a post on Twitter. In doing so the Court considered in detail the scope and operation of injunctions on Twitter, a platform with much of its operations located outside Australia.

FACTS

Between 16 and 19 May the first offending tweets appeared [6] with the author of the tweets used a twitter handle that falsely adopted the name of the plaintiff’s CEO.

On 19 May, the plaintiff’s solicitors wrote to Twitter Inc:

• drawing attention to the tweets,
• the offending information contained in them and
• the user’s impersonation of the plaintiff’s CEO.
• requesting Twitter Inc to:
  • remove the offending material from the Twitter website;
  • to deactivate the ‘fake’ user’s account;
  • to take all other steps available to it to prevent the user from publishing further confidential information on the Twitter website; and
  • to provide the identity and contact information of the user.

Twitter responded promptly, stating that it had ‘removed the reported account for a violation of Twitter Rules specifically the rules regarding impersonation on Twitter [6]

On 30 May, the plaintiff’s solicitors wrote again to Twitter Inc:

• noting that the account from which the offending material had been published had been deactivated,
• adding that the circumstances gave rise to a cause of action against the user for breach of confidence,
• repeating their request for information relating to the identity and contact details of the user [7].

Twitter Inc rejected the request, stating, at [7]:

‘Per our Privacy Policy, Twitter does not release user information except as required by valid legal process’ (emphasis added).

On 13 June, another offending tweet appeared, using a twitter handle also based on an impersonation, but of a different senior officer of the plaintiff. The followers of this tweet were substantially identical to the followers of the first group of tweets [8].

The plaintiff’s solicitors wrote again to Twitter Inc, which again responded promptly stating

‘the account in question has been permanently suspended’

and

 ‘if you would like to request information about the Twitter account and user associated with the suspended account, please complete our Privacy Form’.

In August the tweets continued, with 11 on 27 August, with the user utilising a “..provocative descriptive noun indicative of the nature of the conduct being undertaken” [9]. Some tweets contained precise financial information about the plaintiff that could only be placed by someone with knowledge of its internal affairs.  Other tweets were clearly suggestive of a malicious intention to harm the plaintiff [9].

Again the plaintiff’s solicitors complained. Twitter responded by stating that because there was no impersonation, the tweets did not violate its Terms of Service, a proposition the Court regarded as being “clearly wrong”.

On 3 September there were a further 17 tweets, from the same twitter handle that had been used on 27 August, which were increasingly bold and threatening. On  following day, the plaintiff’s solicitors lodged another complaint but the response from ‘was formulaic and unsatisfactory [12].  The Head of HR and Communications at Twitter Australia  was sympathetic but hamstrung [13]. On 6 September, the plaintiff commenced the proceeding with the court granting  initial interlocutory injunctions including orders that the defendants:

• be restrained from publishing the Offending Material;
• cause it to be removed from the Twitter platform and their web sites;
• suspend the accounts of the user of the three accounts from which the posts had emanated [14].

On 9 and 10 September, a fourth round of offending tweets appeared  adopting a twitter handle that was a variation of the 27 August and 3 September tweets, stating among other things  ‘We are back up!’[15]. Twitter quickly shut down the account.

By 10 September, four separate handles had been used to publish the plaintiff’s confidential information on the Twitter platform [15].

On 15 September the court conducted the final hearing and heard evidence and submissions from the plaintiff’s legal representatives. The defendants were on notice of the hearing and the relief claimed but elected not to appear and made a point of stating that they were not submitting to the jurisdiction [16].

DECISION

The court found there could be no argument about:

• the confidentiality of the financial information disclosed in the tweets.
• the strict obligations of confidence owed to the plaintiff by the person or persons who disclosed that information [5].

The claim for final injunctions against the defendants relating to the unauthorised publication of the plaintiff’s confidential information which was made in the form of ‘tweets’ on  Twitter [1].

The court noted that the defendants are foreign corporations with:

• Twitter Inc being a Delaware corporation whose head office and principal place of business are in California.
• Twitter International Company is an Irish corporation whose principal place of business is in Dublin.

The court framed the substantial issue for determination as being:

• the jurisdiction to grant injunctive relief against foreign defendants who do not appear;
• the appropriateness of injunctions expressed to operate ‘everywhere in the world’;
• whether there should also be orders requiring the defendants to disclose to the plaintiff the name, address, contact details and IP address of the account holder and author of the tweets; and
• whether an appropriate balance between fairness to the plaintiff on the one hand, and the public interest in open justice on the other hand, justifies an order suppressing the identity of the defendants, as well as that of the plaintiff [4]

The court noted that:

• there was no necessity to prove that Twitter was ‘knowingly concerned’ in the user’s breach of duty as against the plaintiff. The cause of action against Twitter was direct & operates independently of the claim against the person originally responsible for the ‘leak’.
• where a third party such as Twitter comes into possession of confidential information and is put on notice of the character of the information and the circumstances in which it was unlawfully obtained, it becomes subject to an equitable obligation of confidence and may be liable to be restrained from publishing the information [17].
• the equitable principle applies whether the recipient of the information is an individual, a newspaper, a magazine, a television channel, a radio station or an online news and social networking platform such as Twitter [19].

The court found it was uncontroversial that the court had jurisdiction as:

• a defendant’s physical presence is not a prerequisite to jurisdiction &
• it is unnecessary to serve the writ on the defendant within the territory of the state as thecourt may allow service out of the state, and by doing so acquire statutory jurisdiction over a foreign defendant [20]
• it can restrain restrain conduct outside Australia. The issue is one of discretion, not power.

In support of those contentions the court cited from its earlier judgement, where it granted interim injunctive relief stating:

•  courts of equity have long exercised jurisdiction to make in personam orders restraining foreign defendants from breaching duties to a plaintiff such as in Australian Competition and Consumer Commission v Chen [2003] FCA 897
• National Australia Bank Ltd v Dessau: which stated that the jurisdiction is not grounded not on any pretension to the exercise of judicial power abroad but on the circumstance that the defendant, being amenable to the Court’s jurisdiction, can be personally directed to act or not to act.

The Court rejected the Defendant’s contention that Macquarie Bank v Berg was authority for the propositions that:

• Courts can only restrain a party to ensure compliance with the laws of NSW’
• a restraint of publication of offending material outside NSW ‘exceeds the proper limits of the use of the injunctive power of the court’.

The court noted the international reach and broad formulation of the proposed final orders sought by the plaintiff  when considering the exercise of the court’s discretion in granting injunctive relief which would impact on Twitter which had over 300 million active users who constantly post an enormous amount of content.  Twitter believed it was not feasible to proactively monitor user content for Offending Material [25]

The original order:

• was to restrain publication of and
• was to require the removal of the ‘Offending Material’,
• was to suspend the relevant accounts from which it emanated.
• defined the Offending Material as:
  • ‘the ‘information contained in or referred to in’ the specified tweets that had emanated from the particular Twitter handles adopted by the person or persons responsible [26].
  • ‘any further tweets posted on the Twitter platform or the defendants’ websites by any person who is the user of one or more of the accounts’ with the same Twitter handle as had been used for the previous tweets… ‘including any new account opened by such a person’ [27].

The plaintiff sought to expand the scope of the proposed final orders to require the defendants, at [29]:

(a) be restrained from publishing the Offending Material anywhere in the world on the Twitter platform, their websites or otherwise;

(b) cause the Offending Material to be removed everywhere in the world from the Twitter platform and their websites;

(c) remove everywhere in the world the accounts held by any person or persons who are the user or users of one or more of the Offending Accounts (namely those used for the offending tweets) ‘to the intent that such person or persons, to the extent known by the defendants, be prevented from operating any account with the defendants’;

(d) be restrained from notifying the users of the Offending Accounts ‘of the removal of the accounts’.

The court stated that in the exercise of discretion regarding injunctive orders:

• should not be ‘exorbitant’ in form or effect,
• should do only the minimum necessary to achieve justice.
• should be formulated ‘with the greatest precision’ so as to make clear what the defendant is required to do or not to do [30].
•  that are excessively narrow may wreak its own injustice’ [31].

The court stated that  a defendant’s right to return to the Court applies whether the injunction is final and perpetual, or merely interlocutory [34].

The Court agreed to make orders substantially in the form proposed by the plaintiff.  It regarded as relevant:

• the malicious intent and apparent determination of the person or persons responsible for the tweets;
• the author(s) had behaved dishonestly and in flagrant breach of his or her duty of confidence;
• the plaintiff’s real concern was the prospect of future tweets of confidential information emerging from the same source, utilising new user accounts and different handles and the defendants were on notice of that prospect [34].
• the court should do what reasonably can be done to prevent any repetition of the damaging tweets that were posted between 16 May and 10 September.
• orders confined to historical information or conduct or existing accounts might be of limited practical utility[34].

The court noted that second part of the definition of Offending Material operated in relation to any future tweets by the user or users responsible for the previous tweets, as well as any new account that may be opened by such a person [27]. It operates prospectively and is unlimited as to time or subject-matter [27]. The court stated that there was no evidence put to the effect that it was not feasible to proactively monitor user content for Offending material  [35].  The court accepted the plaintiff’s submission that there must be a mechanism to filter information on the Twitter service [36] and it was not unreasonable:

• to make orders that require some degree of filtering, or checking, to ensure that the information either does not get posted or, if it is posted, it is removed [36].
• that the proposed orders in relation to future tweets and future accounts are not subject to limitation as to subject matter; or that they apply to any accounts held by any person or persons who are the user or users of one or more of the Offending Accounts [37].

Because of the behaviour of those responsible for the offending tweets to date the court stated it  could not be assumed safely that the content of any future tweets from the same source would be innocuous & in fact the inference was to the contrary.  The orders required an objective decision about the origin of future tweets with the focus being on the person who is the user of one or more’ of the Offending Accounts [37].

The court did not regard proof of the means of ensuring compliance in foreign jurisdictions as a pre-requisite to the grant of the injunctions [38] because:

• ‘Equity acts in personam’, with the proposed orders being a personal direction to perform or abstain from performing particular acts. They do not affect the proprietary rights of the defendants. Courts of equity may make in personam orders that are intended to operate extra-territorially [39].
• the high profile and size of Twitter gives  confidence in compliance as Twitter is a responsible worldwide organisation of significant influence and financial strength which has an undoubted commercial interest in ensuring that Twitter complies with the laws, and the orders of courts, in those countries in which its services are provided [40]
• Twitter’s conduct to date was such that it was likely it would use its best endeavours to give effect to the proposed orders especially as the gist of the orders regarding future tweets and future accounts goes to user identity. As such they do not require ‘pro-active monitoring’ of content [41]
• there was a public interest in making the proposed orders to demonstrate that wrongful conduct will be remedied as effectively as can be achieved [42].
•  Twitter is continually seeking to improve its ability to eliminate spam accounts from the calculation of active users, resulting in the suspension of a large number of accounts that it has established an information security program designed to protect non-public consumer information [43]

The Court made identity disclosure orders, at [48], to facilitate its claim against the person or persons responsible for the  tweets as satisfying the requirements of a Norwhich Order being:

• the party from whom discovery is sought has become ‘mixed up’ in the subject matter for which discovery is required.
• the plaintiff has established that there has been wrong-doing; and third, the plaintiff has established that it needs the discovery in order that it not be denied justice [47].

The court made the orders because, at [48]:

(a) The offending tweets contained the plaintiff’s confidential information. That information is highly commercially sensitive and was required to be kept confidential by the partners and employees of the plaintiff. The disclosure of the information would provide the plaintiff’s competitors with a substantial commercial or competitive advantage over the plaintiff;

(b) The plaintiff’s confidential information has been published through at least four accounts on the Twitter platform and websites. Those publications were made without the knowledge, consent or authorisation of the plaintiff. An inference is available that the user or users of the four accounts is the same;

(c) The plaintiff does not know the identity of the person or person’s responsible for the offending tweets and therefore cannot yet restrain them from further publishing its confidential information. It is unable to commence proceedings against that person or those persons for breach of confidence;

(d) The defendants will have, at least, the name, contact details and IP address of the person or persons who established or held the accounts from which the offending tweets emanated; and

(e) If the person or persons responsible are not themselves restrained from further publishing the plaintiff’s confidential information, there is a foreseeable risk that the plaintiff may suffer significant and irreparable damage.

The orders  were that the defendants provide to the plaintiff’s solicitors, in relation to the accounts with the four separate handles that were used for the offending tweets, and to the extent to which the information is available to or is retained by the defendants:

(a) – the subscriber information for those accounts;

(b) – the name or names of all users of those accounts;

(c) – the IP address or addresses and associated information relating to each of those accounts;

(d) – all phone numbers associated with any user or users of those accounts;

(e) – the location of all users of those accounts and the location of any IP address or addresses associated with those accounts;

(f) – any Twitter or other accounts which ‘are or were at any time’ used by any user of the above accounts or which ‘have or at any time had’ the same phone number or email address as any of those accounts.

The court also made suppression orders in relation to any information that tends to reveal the identity of the plaintiff, [50], as it was in the interests of the administration of justice that the very proceedings before the court should not be permitted to destroy or seriously depreciate the value of such confidential information, [51].  The court did not make any suppression order relating to  identification of the names of the defendants, including the use of the word ‘Twitter’ because:

• it would not be likely to have the same consequence and cause the same prejudice.
• there was a public interest in general awareness of the facts and circumstances of this case
• the public interest in open justice favours transparency as far as possible [53]

The court ordered that costs should follow the event, even if the defendants chose not to appear. The defendant’s liability to costs is the usual, salutary and important consequence of a plaintiff succeeding [54].

ISSUE

Issues of on line identification, Norwhich orders and take down notices are quite established in the defamation practice. This decision highlights that a social media platform such as Twitter can also be used for other forms of civil wrong, here a breach of confidence.  The court went through the process of finalising injunctive relief.  In doing so it carefully considered the operation of those orders  in terms of compliance.   The court was unimpressed with Twitter’s assertion that it was not practical to comply with the orders prospectively without putting any evidence before the court.  It was a bad tactical decision for Twitter not to participate in the proceedings.  That undermines any appeal, if one was ever in prospect.

Arguments that social media platforms are incapable to ferreting out anti social behaviour, illegal comments and Fake News ring increasingly hollow as Governments, commentators and even courts become more and more aware that alogorithms coupled with the application of resources are able to deal with those issues.

The courts will continue to apply current procedures to acts taking place in the cyber sphere.  The principles remain the same.