Algorithmic Laws & Contracts
Xalgorithms undertakes and sponsors research towards ensuring that the concepts, methods, roles, systems and services of an Internet of Rules (IoR) will operate in alignment with legal frameworks across diverse jurisdictions.
Transaction Prerogatives By Jurisdiction
Within each legal jurisdiction throughout the world, buyers, sellers and intermediaries in commerce hold rights and responsibilities in particular ways, underlying their prerogatives when forming contractual relationships. An automated system must obtain information about these factors in a reliable, verifiable way. The first practical requirement is to obtain information about which jurisdictions are involved in any particular transaction. The term “prerogative” refers to pre-eminent authority established in law of any jurisdiction which establishes rights and responsibilities apportioned throughout multi-sector and multi-level society (e.g. citizen–household–community–municipality–province/state–country–multilateral entity).
Since 2012, the Internet & Jurisdiction multi-stakeholder global forum has hosted discussions of both the operational solutions and ethical considerations in reconciling the inherent cross-border nature of the Internet with the prerogatives of national and sub-national jurisdiction. Xalgorithms Foundation monitors and builds upon the I&J community’s work and implements ISO Country Codes (ISO-3166-1) and ISO Region Codes (ISO-3166-2) in order to develop and maintain a computational table of the legal prerogatives of stakeholders in each jurisdiction vis-à-vis the data arising from a transaction.
The reference table of transaction prerogatives would enable an instance of Lichen to contextualize the role-based rights and responsibilities of transaction stakeholders, so that each is informed and empowered to exercise their respective lawful prerogatives without these being inadvertently constrained by the transaction software. Stated the other way around: no stakeholder should be prevented from exercising their full lawful prerogatives by the providers of transaction software.
In instances of apparently overlapping jurisdiction, Xalgorithms cannot anticipate what a court would be expected to rule. The limited pragmatic role of a reference table on transaction prerogatives by jurisdiction consists in relying upon the most authoritative available source(s) of information on transaction-stakeholder prerogatives that appear to be in force in various jursdictions, and the operational components must accurately communicate this information to the parties in a transaction. The operational components of an IoR must provide a straightforward way for the buyer and/or seller to access explanatory detail about each automatically invoked rule. In the case that many rules are invoked all at once, the user interface must not become cluttered or cumbersome. Lichen’s custom-designed navigator is bring optimized to deliver a considerable volume of rules information to users in a scalable way, yet in ‘bite-size’ pieces such that it remains accessible.
An Internet of Rules (IoR) will inevitably implement one of the following positions, with regard to the prerogatives and roles of buyers, sellers, governments and suppliers of transaction-intermediary systems (where the latter includes the providers of IoR components, including the Xalgorithms development team).
- 1. The specification explicitly or implicitly leaves out-of-scope the issue of whether or not IoR components could be enabled to impose and/or constrain the automated incorporation of rules to transactions in accordance with the lawful prerogatives of payers, payees and governments (as documented in the TransactionPrerogativesByJurisdiction table);
- 2. The specification explicitly identifies the criteria, methods and processes by which IoR components would be enabled to impose and/or constrain the automated incorporation of rules to transactions in accordance with the lawful prerogatives of payers, payees and governments;
- 3. The specification explicitly requires, and documents, as to how IoR components must be enabled to impose and/or constrain the automated incorporation of rules to transactions in accordance with the lawful prerogatives of payers, payees and governments;
- 4. The specification explicitly prohibits IoR components from obstructing payers, payees or governments in the exercise of their lawful prerogatives to impose and/or to constrain the attachment of rules to transactions, and any programmatic imposition and/or constraint on the automated incorporation of rules to transactions by the components themselves would violate the specification.
Xalgorithms implements the fourth option above, upon the premise that the Lichen component is not an enforcement method. A buyer or a seller must always retain the straightforward technical ability to remove any rules introduced through Lichen’s auto-generated results. However, if Lichen-generated results are removed by a user, the data in MyLichenCollection would record that the user was informed of, and yet elected to remove one or more particular rules.
Some people may not like that government-revenue agencies, in their capacity as transaction stakeholders, will be further empowered to exercise their lawful prerogatives more efficiently and effectively (to automatically collect taxes, tariffs and other fees). But just as this auxiliary transaction software should not serve as an enforcement method, it should not itself serve as an informal evasion method. Use of transaction software as an ad hoc work-around to evade taxes or tariffs does not provide the depth of legality, sustainability or universality that would be obtained if instead, through political engagement, the same stakeholders were to succeed in changing the tax or tariff laws, or even the legal prerogatives that prevail within their jurisdictions.
Global Legal Frameworks
Xalgorithms monitors a variety of trans-jurisdicational and multi-jurisdictional frameworks that establish the general principles, concepts and definitions of public law and commercial law governing transactions. For example, we attempt to ensure that the systems we implememt align to:
- UNCITRAL WG IV
- “Model Law on Electronic Commerce”
- “United Nations Convention on the Use of Electronic Communications in International Contracts“
- Related UNCITRAL guides and interpretations
- UNIDROIT “Principles of International Commercial Contracts”;
- IOSCO “Objectives and Principles of Securities Regulation”;
- BIS “Principles for Financial Market Infrastructures”;
Ongoing Advances in Legal Research
Xalgorithms monitors relevant academic journals, legislation, case law and the work of specialized institutes for ideas and debates relating to the thoughtful design and deployment of computational algorithms in commerce, the automation of legislation and regulation, and ways that algorithms can enhance rather than erode market competition and integrity. Following are references to some particularly insightful work that we have noticed online:
- Dr. Dag Wiese Schartum of University of Oslo (Faculty of Law).
- Schartum, D. W. (2016a). From Algorithmic Law to Automation-friendly Legislation.
- Schartum, D. W. (2016b). Law and algorithms in the public domain. Etikk I Praksis – Nordic Journal of Applied Ethics, 10(1), 15.
- Computational Law Research and Development at MIT
- Lauren Henry Scholz, Foundations of Private Law Project, Harvard Law School.
- Scholz, L. H. (2016). Algorithmic Contracts. Stanford Technology Law Review, forthcoming.
- In 2016, the OECD produced a report entitled “Bringing Competition Policy to the Digital Era”, which was considerably influenced by a recent book entitled “Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy” by Ariel Ezrachi and Maurice Stucke, former attorneys with the U.S. Department of Justice Antitrust Division. Unfortunately neither Ezrachi and Stucke, nor the OECD report authors, devote adequate attention to exploring ways that algorithmic commerce may enhance competition. Without downplaying the risks that algorithms present to market fairness, arguably there may also be advantages to pursue through algorithmic commerce to make markets more competitive than they would otherwise be. These works are included here because they effectively describe the many risks to competition. We look forward to equivalent attention to many ways that algorithmic commerce can enhance competition.
- Ezrachi, A., & Stucke, M. E. (2016). Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy. Cambridge, Massachusetts: Harvard University Press.
- Stucke, M. & Grunes, A. (2015). Big mistakes regarding big data. University of Tennessee, Knoxville. Research Paper #276.
- OECD. (2016). Big Data: Bringing Competition Policy to the Digital Era. Background note by the Secretariat, DAF/COMP(2016)14, Version dated 2016-10-27. Organisation for Economic Co-operation and Development, Directorate for Financial and Enterprise Affairs, Competition Committee. Retrieved from http://www.oecd.org/daf/competition/big-data-bringing-competition-policy-to-the-digital-era.htm
- Ezrachi and Stucke identify four types of potential collusion:
- Firms that use real-time data to monitor compliance with an explicit agreement on price may be deemed to be intentionally or inadvertently functioning as a conventional cartel.
- Multiple firms that use the same pricing algorithm to simultaneously adjust prices based on market data, may be deemed to be intentionally or inadvertently functioning as a hub-and-spoke cartel.
- Firms that exchange information to enable greater market transparency and to build interdependence amongst their business may be deemed to be intentionally or inadvertently colluding for the purpose of fixing prices.
- Firms using algorithms that ‘learn’ to optimize price based on shared industry data may be deemed to be intentionally or inadvertently colluding.