1. Outsourcing to the cloud: Caveat emptor or caveat venditor
Corinne Rogers
University of British Columbia
XVI Congrés d’Arxivistica de Catalunya
Associació d’Arxivers-Gestors de Documents de Catalunya
May 4-6, 2017
Reus, Catalonia, Spain

2. Agenda Introduce InterPARES
IP1, IP2, IP3, InterPARES Trust
Discuss pros and cons of cloud computing for recordkeeping
“the cloud” in 2017: status, benefits, challenges
Present several tools for evaluating cloud services from the perspective of archival theory
Coping with retention & disposition in the cloud
Evaluating CSP contracts

3. InterPARES – 4 phases to date
InterPARES 1 ( )
Researched issues pertaining to digital records in databases and office management systems in the course of administrative activity
Focused on developing theory and methods to ensure preservation of authenticity
Studied records from the perspective of the records preserver

4. InterPARES – 4 phases to date
InterPARES 2 ( )
Researched issues pertaining to digital records in dynamic and interactive systems in artistic, scientific, and government activity
Examined issues of authenticity, reliability, and accuracy over the lifecycle
Studied records from the perspective of the records creator

5. Benchmark Requirements supporting the presumption of authenticity

6. Baseline Requirements supporting the production of authentic copies

7. Products Creator & Preserver Guidelines

8. Guidelines translated
Catalan
French
Portuguese
Spanish

9. InterPARES – 4 phases to date
InterPARES 3 ( )
Put theory into practice in archives / records units in organizations with limited financial or human resources
Applied and tested the findings of InterPARES 1 and 2 to implement sound programs supporting the creation and preservation of digital records that could be shown to be authentic, reliable, accurate

10. Impact Legislation: Italy, China
Standards: DOD (2007), MoReq 2 (2008), OAIS (2009), CGSB (2017)
Policies & procedures: all participating countries, public/private sector
Curriculum for continuing education, university training: ICA Education Modules for Digital Preservation (2012 with translation to Chinese, Spanish, Arabic); Digital Diplomatics and Digital Records Forensics (2013-present, UBC)

11. InterPARES Trust (2013-2018) www.interparestrust.org Purpose:
To generate theoretical & methodological frameworks to support development of integrated & consistent local, national, & international networks of policies, procedures, regulations, standards, & legislation for digital records in online, networked, environments, in order to
Ensure public trust grounded on evidence of good governance, strong digital economy, & persistent digital memory

12. Research structure
Studies are focused in 5 research domains and 5 research cross-domains:
Access
Control
Security
Infrastructure
Legal issues
Policy
Social issues
Terminology
Resources
Education

13. Research structure
InterPARES Trust is a research partnership led by UBC:
National libraries and archives
Government departments: national, regional, municipal
Academic departments
International organizations
Private industry
Not-for-profit consortia

14. Research structure
Partners are organized in regional teams spanning 6 continents
North American Team
Latin American Team
European Team
Asian Team
African Team
Australasian Team
Transnational Team

15. What are we researching?
What is the impact of always-on, networked communications technologies and cloud computing services on records management & recordkeeping, maintaining trustworthy records & supporting client/citizen perception of trustworthiness of records?

16. Goals
To discover how current policies and practices regarding the handling of digital records in online environments by institutions and professionals affect public trust
In other words, what are records professionals doing to maintain trustworthy records?

17. Goals
To anticipate problems in maintaining trust in digital records under the control of entities currently suffering a waning level of confidence from the public
In other words, what is the public’s perception of the trustworthiness of institutional records?

18. Goals
To establish what significance national or cultural contexts have with regards to the level of trust in digital records online
To develop model policies, procedures, guidelines, standards, & functional requirements for creating, managing, accessing, storing, preserving trustworthy records online
To test these instruments in a variety of contexts

19. Why is this research necessary?
“Cloud-first strategies are the foundation for staying relevant in a fast-paced world” Gartner, 2015
“Enterprise adoption of the cloud has truly moved into the mainstream, with 68% currently using public or private cloud… a 61% increase over last year…” IDC, 2016
“The greater the level of cloud adoption, the higher the level of business benefits achieved” IDC, 2016
“On average, per application deployed on cloud, organizations studied are achieving $3 million in additional revenue… [and] $1 million in cost reduction…” IDC, 2016
Enticing, but…

20. What is ‘the cloud’?
“There is no cloud. It’s just someone else’s computer.” (Popular)
True?
“… if you’re saying that, the joke is on you, because it means you don’t understand what the cloud actually is.” (Branscombe, 2017)

21. Cloud computing: NIST
“A model for enabling convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction.”
Delivered in one or a combination of deployment models: public, private, community, hybrid
Three main service models: SaaS, PaaS, IaaS

22. NIST interpreted
“The simplest definition of cloud is a data centre that’s full of identical hardware … [in which] every deployment, update, investigation, and management process is automated.” Branscombe, 2017

23. Challenges
Most challenges discussed represent present concerns with current data (data-centric thinking):
Is data secure from alteration or interference?
Can personal privacy be protected?
Can regulations and laws be observed in the face of cross-jurisdictional data transfer?
What guarantees of continuity of service exist?
How will data breaches be handled?

24. Chain of custody
We keep records (sometimes over long periods of time) as evidence of activity, and as memory of action, & to prove accountability – we must trust them
In archival terms, we trust records based on proof of records’ authenticity, reliability, & accuracy
In legal terms, trust is expressed through rules of admissibility of documentary evidence (common law systems)
Demonstrable chain of responsible custody is key to both

25. Challenges
Recordkeeping challenges look beyond the immediate present, reaching into the past, and projecting into the future (record-centric thinking)
Can context of records be protected?
Can provenance be demonstrated?
Can retention & disposition be carried out?
Can access and usability be assured over time?
Can intellectual rights be respected?

26. Trust
N. 1. Confidence of one party in another, based on alignment of value systems with respect to specific actions or benefits, and involving a relationship of voluntary vulnerability, dependence, and reliance, based on risk assessment.
V. 2. To have confidence in another party with respect to specific actions or benefits
Trust is subjective, existing on a continuum from trust to skepticism

27. Record Trustworthiness
Trust framework
Record Trustworthiness
Authenticity
Identity
Integrity
Reliability
Completeness
Control
Accuracy
Precision

28. Trustworthy records systems
These records requirements depend on trustworthy, controlled systems
Do cloud services meet the standard of trustworthy records systems?

29. Trustworthy records systems: Managing records wherever they are
Whether managing records in a paper-based in-house system, or managing any valued organizational asset, a management framework consists of:
Laws & policies establishing accountability
Standards & practices for management
Systems & technologies for implementation
People
Organizational structure
Awareness & continuing education
Managing Records of Citizen Engagement Initiatives: A Primer

30. Hierarchy of contexts Juridical/Administrative Provenancial Procedural
Documentary
Technological

31. Holistic view of considerations for adopting cloud services
Managerial, including Records Management
Economic
Legal
Security
Technical
Records in the Cloud – Switzerland (2016)

32. Cloud maturity Ad hoc Opportunistic Repeatable Managed Optimized
IDC, 2016

33. Tools for evaluation
Regardless of the degree of cloud adoption, there are tools to evaluate the benefits and risks from the perspective of recordkeeping based on archival science
Checklist for evaluating cloud service provider contracts
Checklist for evaluating retention & disposition capacity

34. CSP contracts as instruments of trust: Purpose & Research question
To explore the contract – specifically the contract between a client and a cloud service provider – as a tool for building trust
How effectively do cloud service contracts meet the needs of records managers, archivists, and information governance professionals?

35. Selected contracts No marketing material
Boilerplate contracts & documents
Terms of Service (ToS)
Service Level Agreements (SLA)
Privacy policies, Acceptable Use policies, Security terms,
Jurisdiction
Canada, United States, Europe
Amazon.com (USA); Bluelock (USA); Dropbox (USA); Egnyte (USA); GoGrid (USA); Google (USA); ProfitBricks (USA); Rackspace (USA); CityNetwork (Sweden); SAP (Belgium); Pathway Communications (Canada)

36. Contracts review Findings: Several legal documents exist
Terms of Service
Service Level Agreements
Privacy Policies
Acceptable Use Policies
Little standardization of terms
“Often incomprehensible to majority of users”
Wide-ranging exclusions of liability favor the providers
Terms may change

37. Related work
Recordkeeping Standards, Cloud Computing Contract Standards, and related articles
Public Records Office of Victoria (2012)
European Commission subgroup on service level agreements (established 2013)
ISO/IEC (2016) SLA Standardization Guidelines

38. CSP contracts in the courts
Case Law and Related Articles
Relatively few cases decided, but several legal tenets involved
Complexity results from jurisdictional and industry differences
Contract law
Privacy and access
Confidentiality and security of data
Data jurisdiction and conflict of laws

39. Comparative Analysis
Regardless of jurisdiction, sector, or industry, common risks to records exist:
Unauthorized access
Privacy breach
Loss of access, control
Lack of transparency of service
Lack of ability to negotiate service
Location ambiguity
Contract ambiguity

40. Specific Considerations
Data ownership
Availability, retrieval and use
Data storage and preservation
Data retention and disposition
Security, confidentiality, privacy
Data location and cross-border data flow
End of service; contract termination

41. The Checklist - sections
Agreement
Data Ownership and Use
Availability, Retrieval, and Use
Data Storage and Preservation
Data Retention and Disposition
Security, Confidentiality, and Privacy
Data Localization and Cross-border Data Flows
End of Service; Contract Termination

42. The Checklist

43. Integration & Review Integrated with NA03: Standards of Practice
Integrated with NA06: Retention & Disposition checklist
Released for comments in fall 2015
Presented at ICA in Rekjavik, Iceland
Tested in several venues including the International Federation of Red Cross and Red Crescent Societies

44. Resources
Cloud Service Contracts: An Issue of Trust, Canadian Journal of Library and Information Science (CJLIS): Special Issue on Data, Records and Archives in the Cloud, June 2015
/Dissemination
Annotated bibliography
Checklist
Final Report

45. Retention & disposition checklist
How does the use of cloud services affect retention & disposition of records in accordance with the law and other applicable guidelines?
Study carried out as part of InterPARES Trust by researchers from San Jose State University (California), British Columbia Government Records Service, archivists & records managers from Universities of BC and Victoria

46. Findings Survey of members of ARMA International: 168 respondents
62% worked in government
60% used some aspect of cloud computing
92% confirmed their organization has a retention policy
50% confirmed that the policy applied to records in cloud storage
69% said that vendor terms and conditions were not consistent with their policies, or they did not know
81% said dispositions on cloud content had not yet been performed, or they did not know

47. Internal & external obstacles
External factors are risk related, or imposed
Internal factors reveal level of cloud maturity knowledge
Differences in IT and RIM culture
Decisions often cost-driven, or made solely by IT department
Lack of knowledge about cloud computing

48. Retention & disposition: questions for evaluation of service
Privacy and security
Establishing disposition authorities
Applying disposition authorities
Executing disposition authorities
Documenting disposal actions
Reviewing disposition
System integration

49. More resources for decision-making
Ensuring Trust in IaaS at
This Checklist is designed to offer guidance for individuals, businesses, government agencies or other organizations to assess the security and ongoing trustworthiness (i.e. authenticity, reliability, and accuracy) of their data when stored in an Infrastructure-as-a- Service (IaaS) platform. It is the result of a study in the international InterPARES Trust Research Project ( Ensuring Trust in Storage in Infrastructure- as-a-Service (EU08). The goal of the study was to establish the minimum amount of information necessary to support users’ trust in an IaaS provider and also position the provider as a trusted service provider.
The checklist consists of 36 questions divided into 10 categories:
General information (4 questions),
Governance (4 questions),
Compliance (4 questions),
Trust (5 questions),
Architecture (6 question),
Identity and Access Management (1 question),
Software Isolation (2 questions),
Data Protection (5 questions),
Availability (2 questions),
Incident Response (3 questions).
This checklist can be used by records managers and archivists when assessing a CSP offering IaaS as well as by CSPs as a guideline for providing online information about their service. The full report from this study can be found at rageIaaS_FinalReport_Final.pdf.

50. Who is responsible? Caveat emptor, or caveat venditor?
Should you outsource IT to the cloud?
Guidance from IDC, 2016
“Simply adopting cloud is not enough; you should increase your cloud maturity level”
“Go with a provider you trust”
Standards of trustworthiness may be identified in terms of levels of expectation of responsibility. The highest level is the fiduciary relationship – that legal relationship between a person in a position of some vulnerability, the trustee, and the trustor, whose aid, advice, or protection is sought on the basis of its justifiable trustworthiness. These phrases express the trust relationship between two parties, where caveat emptor is the level of trust with the least expectation of responsibility – this is the type of trust represented in the ordinary marketplace – buyer beware. But the principle of caveat emptor only works when the buyer has sufficient knowledge of the product he is buying and opportunity for inspection. This was the way of medieval fairs, but doesn’t serve the consumer in the digital age. The mirror image of caveat emptor is caveat venditor – let the seller beware – where the responsibility is shifted to the seller.

51. Who is responsible? Caveat emptor, or caveat venditor?
Selected tools to help, InterPARES Trust, 2017
Checklist for ensuring trust in SaaS (EN, SP)
Checklist for comparative analysis of governmental e-services
Checklist for single sign-on systems
Economic models for could storage decision-making
Archival standard of practice
Functional requirements for retention & disposition in cloud
Managing records of citizen engagement initiatives: a primer
Checklist for evaluating cloud contracts (EN, AP, FR, NL)

52. www.interparestrust.org www.interparestrust.com corinne.rogers@ubc.ca
Tag cloud by Ashashyou (Own work) [CC BY-SA 4.0 ( via Wikimedia Commons