Cybersurveillance Developments

Over the past few months there has been a flurry of sometimes contradictory activity concerning the government’s ability to access electronic information in the course of a criminal investigation.  This article highlights three recent proposals that show how the broader policy debate is playing out at the level of specific legal rules.

Changes to the Federal Rules of Criminal Procedure Concerning Search Warrants

On April 28, 2016, the Supreme Court adopted changes to F. R. Crim. Pro. 41, adding a subsection (6), to authorize a magistrate judge in any district “where activities related to a crime may have occurred” to issue a warrant “to use remote access to search electronic storage media and to seize or copy electronically stored information located within or outside that district. . . .”  See Rule changes submitted by Justice John G. Roberts to Congress, April 28, 2016.  Under the amendment, such warrants can issue if “the district where the media or information is located has been concealed through technological means” or in cases involving investigations of hacking or malware transmission under the Computer Fraud and Abuse Act where the “media” are damaged computers in five or more districts.  Id., amended F. R. Crim. Pro. 41(6)(A), (B).

Previously, the general principle was that a warrant could only be issued to search and seize a person or property located outside the district “if the person or property is located within the district when the warrant is issued but might move or be moved outside the district before the warrant is executed.”  Fed. R. Crim. P. 41(b)(1)-(2).   This principle previously was expanded to include authority to issue a warrant for a person or property outside the district if the investigation involved domestic or international terrorism and to include warrants for installation of a tracking device to track the movement of person inside or outside the district. See Fed. R. Crim. P. 41(b)(1)-(4).  Finally, historically a warrant could be issued for property outside the district but within a U.S. territory, possession or commonwealth, on the premises of a U.S. diplomatic or consular mission in a foreign state, or in a residence leased by the U.S. and used by U.S. personnel assigned to a U.S. diplomatic or consular mission in a foreign state.  Fed. R. Crim. P. 41(5).  

Critics of the recent addition of subsection (6), including some tech industry giants such as Google Inc., argued that “remote access” warrants for nationwide or even worldwide electronic surveillance.  See Google Public Policy Blog, “A Small Rule Change that Could Give the U.S. Government Sweeping New Warrant Power.”  Google’s comments in this regard were typical of tech industry concerns:  

The proposed change does not define what a “remote search” is or under what circumstances and conditions a remote search can be undertaken; it merely assumes such searches, whatever they may be, are constitutional and otherwise legal.  It carries with it the specter of government hacking without any Congressional debate or democratic policymaking process.

Id.  Notwithstanding such objections, the Rule change was approved by the Supreme Court, and will become effective unless disavowed by Congress before December 1, 2016 under the Rules Enabling Act.  See 28 U.S.C. § 2074.

Burr-Feinstein Bill

On April 13, 2016, Senators Richard Burr (R-N.C.) and Diane Feinstein (D-Calif.), Chair and Vice-Chair, respectively, of the Senate Intelligence Committee, released a draft Bill titled the “Compliance With Court Orders Act of 2016.”  See April 13, 2016 Press Release; Discussion Draft.  This Bill responds to the recent showdowns between Apple Inc. and the FBi concerning the ability to compel technology companies under the All Writs Act to assist with access to locked and encrypted devices such as iPhones.  See David W. Opderbeck, “The Apple iPhone Showdown:  What is at Stake,” New Jersey Law Journal, March 7, 2016.  The Bill would require any covered entities that receive court orders “for information or data” to provide the information or data “in an intelligible format” and to “provide such technical assistance as is necessary to obtain such information or data in an intelligible format or to achieve the purpose of the court order.”  Discussion Draft, Sec. 3(a)(1).  The Bill states that a covered entity is only responsible for providing data in an intelligible format “if such data has been made unintelligible by a feature, product, or service owned, controlled, created, or provided, by the covered entity or a by a third party on behalf of the covered entity.”  Id., Sec. 3(a)(2).  The Bill further states that it would not authorize any government officer to require or prohibit “any specific design or operating system to be adopted.”  Id., Sec. 3(b).  However, the very next subsection of the Bill requires providers of “remote computing service” or “electronic communication service” to ensure that their products or services a capable of complying with the requirement to provide data in an intelligible format.  Id., Sec. 3(d), (e).  The terms “remote computing service” and “electronic communication services” are defined to have the meanings provided in the Electronic Communication Privacy Act (ECPA), 18 U.S.C. s 2510, 2711.

The draft Bill was immediately pilloried by technology industry and civil liberties advocates.  For example, Kevin Bankston, Director of the New America Foundation’s Open Technology Institute, called it “easily the most ludicrous, dangerous, technically illiterate proposal I’ve ever seen.”  Andy Greenberg, “The Senate’s Draft Encryption Bill is Ludicrous, Dangerous, Technically Illiterate,” Wired Security, April 8, 2016.  Critics noted that the Bill’s performance standard necessarily would constrain design choices, that it would effectively outlaw user-directed end-to-end encryption, and that it would require a greater level of technological assistance than the government ever sought in the All Writs Act cases.  See “The Burr-Feinstein Proposal is Simply Anti-Security,” Electronic Frontier Foundation Deeplinks Blog, April 8, 2016.

Proposed Amendments to ECPA

The changes to F. R. Crim. P. 41 and the Burr-Feinstein Bill are pro-law-enforcement and anti-encryption.  Not all recent legislative proposals, however, fall on that side of the line.  On April 27, 2016, the “Email Privacy Act” passed the House of Representatives.  See H.R. 699, 114th Cong. 2d Sess. (2015-2016).  The Email Privacy Act would amend the ECPA to require the government to obtain a search warrant to access stored electronic communications.  

The law makes a distinction between electronic communications in transit and in storage.  For communications in transit, the Wiretap Act requires a showing of probable cause plus a showing that “normal investigative procedures have been tried and have failed or reasonably appear to be unlikely to succeed if tried or to be too dangerous.”  18 U.S.C. s 2518(3).  Wiretap orders must expire after thirty days, although extensions are possible upon a showing of necessity.  Id. s. 2518(5).  For communications in storage, presently, the ECPA distinguishes between contents stored by an “electronic communication service (ECS)” and a “remote computing service (RCS),” and as to an ECS, further distinguishes whether the communications have been in storage for 180 days or more.  See 18 U.S.C. 2703.  Finally, the ECPA allows a judge in any district, not only the district where the information is stored, to issue the order.  Id. s. 2703(d).  

Uunder the ECPA, to obtain the contents of stored electronic communications (such as emails and voicemails ) that have been in storage by an ECS for 180 days or less by obtaining a warrant.  18 U.S.C. s 2703(a).  However, The government may obtain the contents of information held by an RCS “solely for the purpose of providing storage or computer processing services,” or held in storage by an ECS for 180 days or more, through a court order based on “specific and articulable facts showing that there are reasonable grounds to believe that the contents of a wire or electronic communication, or the records or other information sought, are relevant and material to an ongoing criminal investigation.”  18 U.S.C. s. 2703(a)-(d).   In other words, the law currently recognizes a lower expectation of privacy (a) for the contents of communications held in storage by an RCS; and (b) for the contents of emails and other communications held in storage for more than 180 days by an ECS.  These distinctions date back to the early days of the Internet, when users were able to download and store only a small amount of data from email servers run by their service providers.  See H. Rept. 114-528 – 114th Congress (2015-2016) April 26, 2016, As Reported by the Judiciary Committee.  

The Email Privacy Act would instead recognize the same expectation of privacy in all communications stored by third party providers by requiring a warrant on probable cause before the government could obtain the contents of such communications, regardless of how long they have been in storage, and regarldess of whether the provider is classified as an RCS or ECS.  See Email Privacy Act, Sec. 3.  This would make the statute consistent with practice in the Sixth Circuit, which has held the distinctions under the present ECPA unconstitutional under the Fourth Amendment.  See United States v. Warshak, 631 F.3d 266 (6th Cir. 2010).  The Bill would not affect the government’s ability to obtain non-content information, such as subscriber records, through an administrative subpoena, nor would it change the ability the owner of a communication system, such as an employer-owned email system, to disclose stored information voluntarily.

Most recently, law enforcement groups sought amendments to the proposed Email Privacy Act, which has stalled the bill’s progress in the Senate.  It is unlikely that any further action will be taken before the Presidential election.

Conclusion

These three recent proposals get “into the weeds” of the larger national policy debate about encryption and Internet surveillance.  They demonstrate that the larger debate implicates a host of more granular authorities involving the scope and requirements of judicially approved process for the government to obtain electronic information and for technology companies to assist with such process.  The critics may be right to worry about the jurisdictional and technological breadth of the changes to the search warrant rule and in the Burr-Feinstein Bill.  However, even if these rules are not adopted and the pro-privacy changes of the Email Privacy Act are enacted into law, significant issues will remain concerning how law enforcement can execute its mission to provide security for everyone while respecting Constitutional privacy concerns in the Internet age.

Managing Cyber Risk: Insurance and Coverage Cases

The scope of the risk and uncertainty involved in data breaches and other cybersecurity incidents suggests that insurance should play a key role in any organization’s risk management strategy. The specialty cyber risk insurance market is rapidly developing, although not yet mature. Among other ambiguities, there is relatively little case law interpreting either traditional or specialty liability policies in connection with cyber liability claims.

The question of how courts will decide cyber insurance litigation became front page business news when a case stemming from the massive Sony Playstation data breach was decided by a New York trial court. Zurich American Insurance v. Sony Corp. of America, No. 651982/2011 (N.Y. Sup. Ct. 2014). The court held that Sony had no coverage for its data breach-related losses under a standard form of Comprehensive General Liability (CGL) policy. In a bench ruling, the court found that there was no “publication” of the data because Sony had tried to keep it secure and it was disclosed only because of the criminal activity of hackers. Sony filed an appeal, but the case settled after appellate arguments.

In another example under a traditional policy, Recall Total Information Management, Inc. v Federal Insurance Co., 83 A.3d 664, 147 Conn. App. 450 (2014), aff’d, 115 A.3d 458, 317 Conn. 46 (2015), a cart full of computer backup tapes containing employee information fell out of a van used by a subcontractor of a records storage company. Some of the tapes were taken by an unknown person and were never recovered. The subcontractor’s insurance policy defined covered “personal injury” to include “injury, other than bodily injury property damage, or advertising injury, caused by an offense of . . . electronic, oral, written or other publication . . . of material that . . . violates a person’s right to privacy.” Id. at 672, 147 Conn. App. at 462. As in the Sony case, the court held that the information on the tapes had not been subject to “publication” because there was no evidence that any third party had actually accessed the information. In addition, the court held that losses relating to notification costs under state data breach notification statutes did not comprise the sort of “injury” covered by the policy. The court therefore granted summary judgment in favor of the insurer. Id. at 671-73, 147 Conn. App. at 461-65.

Not every case, however, has been decided in favor of the insurer. A good example of a case involving a traditional liability policy in which insured prevailed is Travelers Indemnity Co v. Portal Healthcare Solutions, LLC, 35 F. Supp. 3d 765 (E.D. Va. 2014). The policy in this case obligated Travelers to cover losses from damages arising from “electronic publication” of material that “gives unreasonable publicity to” or “discloses information about” a person’s “private life.” Id. at 767. Portal Healthcare provided electronic medical records services to Glen Falls Hospital. Due to an unspecified error, which did not seem to involve hacking, two patients of Glen Falls who searched their own names on Google discovered that their medical records came up as the first search result. Id. at 768. The court held that the records were subject to “publication” because they were made available over the Internet, even though their availability was unintentional and even though there was no evidence they had been accessed by any member of the general public. The court also found that this constituted unreasonable publicity and disclosure as those terms were used in the policies. Therefore, the court granted summary judgment in favor of the insured. Id. at 771-72.

There will undoubtedly continue to be claims and litigation over traditional policies for some time. However, the Insurance Services Office’s (“ISO”) standard commercial general liability (“CGL”) policy forms were changed by ISO in 2013 and 2014 to limit and exclude coverage for privacy and data breach claims. See CG 24 13 04 13 (2012) (limiting personal and advertising injury liability); CG 21 06 05 14 – Exclusion for Access or Disclosure of Confidential or Personal Information and Data-Related Liability – With Bodily Injury Exception; CG 21 07 05 14 – Exclusion for Access or Disclosure of Confidential or Personal Information and Data-Related Liability – Limited Bodily Injury Exception Not Included; CG 21 08 05 14 – Exclusion for Access or Disclosure of Confidential or Personal Information (Coverage B Only).

The specialty cyber risk market will therefore become increasingly important. Cases are just beginning to arise under these specialty policies.  A recent case in federal district court in Utah has been described as the first coverage case involving a cyber risk coverage form. See Travelers v. Federal Recovery Services, Inc., 103 F. Supp.1297 (D. Utah 2015). The relevant form was a Travelers “Cyberfirst Policy” with a “Technology Errors and Omissions Liability Form” which provided coverage for an “errors and omissions wrongful act.” Id. at 1298. “Errors and omissions wrongful act” was defined in the policy as “any error, omission, or negligent act.” Id. at 1298-99.

Federal Recovery Acceptance (“FRA”), a data storage, processing, and backup company, had a services contract to handle a client’s customer account data. The client was being acquired by another company and requested return of its client data. FRA allegedly withheld the data and demanded a ransom payment beyond that allowed under its contract. The client sued FRA and related defendants for tortious interference, promissory estoppel, conversion, breach of contract, and breach of the implied covenant of good faith and fair dealing. FRA and the related defendants tendered the defense to Travelers, which filed an action for declaratory relief and accepted the tender under a full reservation of rights. Id. at 1299-1300.

The court held that Travelers did not have a duty to defend because “none of Global’s allegations involve errors, omissions, or negligence.” Id. at 1302. Instead, FRA was allegedly intentionally withholding the data.
Another case involving a specialty cyber risk policy recently was recently filed in federal court in California. Columbia Casualty Co. v. Cottage Health System, No. 2:15-cv-03432 (C.D. Cal.). Columbia had issued a “NetProtect 360” policy to Cottage Health. See Complaint for Declaratory Judgement and Reimbursement of Defense and Settlement Payments (“Complaint”), Columbia Casualty Co. v. Cottage Health System, Case No. 2:1-cv-03432 (C.D. Cal. May 7, 2015), 2015 WL 2201797. The policy covered losses for “privacy injury” and “privacy regulation proceedings.” However, the policy contained an exclusion for “failure to follow minimum required practices.” The exclusion stated that Columbia is not liable for any loss “based upon, directly or indirectly arising out of, or in any way involving . . . any failure of an Insured to continuously implement the procedures and risk controls identified in the Insured’s application for this insurance and all related information submitted to the Insurer in conjunction with such application . . . .” Complaint, ¶ 26. As part of its application for the policy, Cottage Health had submitted a “Risk Control Assessment,” which required it to respond to questions about security patches, threat assessment, audits of third party vendors, and other cybersecurity practices. Id., ¶ 29.

Cottage Health System’s electronic medical records provider allegedly failed to secure a server with encryption, resulting in the exposure of approximately 32,500 patient records. Cottage Health was named in a class action under California’s Confidentiality of Medical Information Act, Cal. Civil Code § 56 et seq., and ultimately agreed to a $4.124 million class settlement. Cottage Health also faced an investigation by the California Department of Justice. Columbia agreed to fund the settlement under a reservation of rights. Complaint, ¶¶15-22. The Complaint alleged, on information and belief, that Cottage Health provided false responses to the security-related questions in the application, and that as a result, Columbia had no liability under the Policy. Complaint, ¶¶30, 39-58.
Columbia’s Complaint was dismissed without prejudice pending the completion of an alternative dispute resolution procedure agreed to by the parties. See Order Granting Motion to Dismiss, Columbia Cas. Co. v. Cottage Health Sys. 2015 WL 4497730 (C.D. Cal. July 17, 2015). Nevertheless, the initiation of this proceeding by Columbia signaled that issuers of cyber risk policies may be willing to test insureds’ compliance with required security programs in court.

If there is a pattern in these cases, it may be that – not surprisingly – insurers will contest coverage for a variety of reasons under both traditional CGL and specialty cyber policies. Indeed, a number of very recently filed coverage cases confirm this intuition. See, e.g., New Hotel Monteleone, LLC v. Certain Underwriters at Lloyd’s of London, No. 15-11711, 2015 WL 9608250 (Orleans Parish, LA, December 10, 2015) (coverage claim arising from hacking incident resulting in stolen payment card information); Certain Underwriters at Lloyd’s of London v. Wunderland Group, LLC, No. 2015-CH-18139, 2015 WL 9608250 (Cook County, Ill., December 15, 2015) (coverage claims relating to insider attack / trade secret misappropriation); Ameriforge Group, Inc. v. Federal Ins. Co., No. 2016-00197, 2016 WL 1366311 (Harris County Tex. January 4, 2016) (coverage claims relating to phishing attack). Even specialty policies may not cover risks such as cyber-extortion by a vendor, or they may require the insured to implement a rigorous cyber hygiene program. The precise language of the coverage and exclusions, as well as the precise nature of the incident, will be at issue in any coverage challenge. An insurance program should be carefully evaluated and cyber hygiene should be continuously monitored as part of a comprehensive cyber risk mitigation strategy.