Corruption

The Western Australia Corruption and Crime Commission's Report into misconduct risks with access to confidential information in the Office of the Auditor General comments

The Auditor General is an independent officer of Parliament and the Auditor General Act 2006 gives the Auditor General independence in decision-making. The Office of the Auditor General (OAG) is a major integrity agency. In carrying out its functions, necessarily the OAG must examine material within public authorities that is confidential and sensitive. For this reason, officers of the OAG are required to keep material they uncover during the course of their work as auditors confidential on pain of penalty. 

On 27 February 2019, as required by law, OAG notified allegations of serious misconduct to the Commission. The misconduct risk in relation to data and information is one of the Commission's strategic themes informing its decisions on possible investigations. In view of the potential misconduct risk in the exposure of confidential information, the Commission commenced an investigation code named Operation Phoenix. 

Although OAG had conducted a preliminary investigation, the seriousness of the allegations were best suited to a Commission investigation which could utilise digital forensic expertise and hold private examinations. 

The outcome is startling. Two auditors, each a certified practicing accountant (CPA), had routinely accessed confidential information about other OAG officers, including payroll details and other private and confidential information. 

They were able to access confidential information within OAG because it was not properly protected. Once an officer in OAG logged on using a password, that officer had access to all of the OAG systems, including access to TRIM, a record management system. Each primary file on TRIM had its own access controls set by OAG staff. The primary file was a dataset in that it contained numerous individual documents relating to that subject. If the officer's staff profile had access to a particular file, that access allowed the officer not only to browse or read the documents contained within that file, but to download the document or migrate it to a private device. 

In its response to a draft of this report, OAG detailed the security requirements and the defect that enabled the breach:

OAG's systems and datasets have additional security requirements on a role-based or need-to-access basis. The payroll system and the finance system for example have strong access controls that limit access only to specifically identified staff. These are not open to all staff. Audit files are only accessible by audit staff in the relevant division, with sensitive audit files having restrictive access to only the relevant audit team. Some (but not all) sensitive/confidential files in TRIM had lost their tight security when new versions were created and the new files did not inherit the access restrictions of the parent folder. This was subsequently addressed with enduring monitoring controls in place.

The officers' conduct described in this report demonstrates that information on the systems such as payroll reports and credit card statements were open to all staff, albeit due to inadequate access controls on TRIM. 

In most organisations, an officer is given access only to those matters required for their role. Access to payroll and personal details of staff is generally confined to members of human resources and finance teams. 

The potential for others to acquire this information is a serious misconduct risk. It can be used for personal gain. It can be sold to criminals. 

In the course of Operation Phoenix, the Commission uncovered a further misconduct risk. 

Auditors from OAG working in teams routinely make site visits to public authorities to conduct audits. 

Security of data provided by the public authority for the purpose of an audit is controlled by use of an encrypted USB flash drive, known as an IronKey. The IronKey is used with a laptop computer provided to each auditor by OAG. 

OAG's policies require information received on an IronKey to be deleted after it is uploaded to the OAG audit program. OAG provides periodic reminders regarding the obligation for staff when information is particularly sensitive. However, if these policies are not followed, information may remain on a laptop for years, able to be copied and shared. 

OAG, in its response to this paragraph said:

OAG has a policy that requires that information received from client agencies is deleted from the IronKey after it is loaded into the OAG's protected audit workpaper system. Further, for the audit in question the Assistant Auditor General gave written instructions to all auditors, reminding them of this and other security requirements around data handling. This was supplemented by verbal advice. The officer in question was one of those given these reminder instructions ... We have also undertaken periodic checks that staff do not retain copies of audited entity information on their laptops and IronKeys

Clearly, whatever policies and instructions were in place, did not work. This remains a serious misconduct risk. 

An auditor obtained and retained access to the names and addresses of every serving police officer in WA, some years after completing an audit of the WA Police Force. The names of 8,800 officers, employees and contractors were stored on a spreadsheet on a laptop computer. OAG was unaware until the laptop was forensically examined as part of Operation Phoenix. There is no evidence the police data was shared with others. However, the misconduct risk is obvious. The information was less than five years old. Its value to criminal elements could be immense. 

OAG has independence of action and is responsible for auditing the finances and actions of all departments of government, State and local. It should be trusted to keep information confidential. The misconduct risk exposed in this report shows, unless OAG has taken action to tighten its controls, that trust may be misplaced. 

OAG, in its response said 'OAG took remedial action immediately following the incident. It is of high priority to the Office that audited entities can place trust in the OAG. Strong information security and continuous improvement is of upmost importance to the OAG'.

Expression and Pseudolegal Absurdity

'Special, Personal and Broad Expression: Exploring Freedom of Expression Norms under the General Data Protection Regulation' by David Erdos comments

The interface between data protection and freedom of expression is increasingly crucial and the GDPR solidifies bipartite or even tripartite conceptualisation of this relationship. Whilst the GDPR’s personal exemption can play some in governing individual expression, it must be construed narrowly so as to only exclude innocuous publication that is not liable to infringe other’s fundamental rights. The special expression derogation remains central and encompasses not just journalism but also other forms of special expression (academic, artistic, literary) which, when published, are objectively orientated towards a collective public. Whilst Member States retain considerable discretion, a strict balancing between fundamental rights must be ensured. Freedom of expression is also distinctly furthered by inter alia self-expression on social networking sites and the facilitation of a range of expressive purposes by search engines. As shown in GC and Others v CNIL, the reconciliation of rights here must retain a direct role for data protection’s core substance including its legal grounds and principles.

Yet another instance of pseudo legalism in Tasmania, with the Court in Police v Gale [2020] TASMC 11 taking a dim view of Gale's claims. It states 

 Despite a number of the complaints involving serious offences and the majority having a protracted history before the court Mr Gale is yet to enter a plea to any of the charges. He has instead raised a variety of preliminary arguments which challenge, inter alia, the legal basis of the matters of complaint, the authority of Tasmania Police to bring such charges and the jurisdiction of the Court. The defendant was provided with the opportunity to substantiate his arguments. In determining the points raised I have had regard to the voluminous materials relied upon by Mr Gale and presented, somewhat erroneously, as an “affidavit” dated 17 October 2019 and an “addendum” dated 18 November 2019. 

The Prosecutor chose not to file any responding materials but rather relied upon Nibbs v Devonport City Council [2015] TASSC 34. I have considered all of the material supplied by the defendant, together with his oral submissions. I will attempt to identify and deal with each point in turn. 

Onwards! 

 I have done my best to decipher the almost unintelligible mixture of archaic jargon and nonsensical sentences set out in the “Certificate” provided by the defendant. I have also reviewed, as best I am able, the mixture of historical and current legislation relied upon. I am mindful of a Judicial Officer’s duty to provide adequate reasons for a decision, not only so that each party can understand the determination and its impact upon their respective interests, but also to allow for review by a superior court. However, in the present situation I am at a loss. To provide such reasons would mean that I could make some sense of the documents and arguments presented. I cannot. As such, all I can do is dismiss the ground of argument set out at paragraph 8(i) and Exhibit 01 of the defendant’s “affidavit”. 

The Court went on to adopt the findings by Magistrate A R McKee in Police v Nibbs [2018] TASMC regarding the validity of Commonweraklth and state law. In Gale it states In reliance upon the case of Hale v Henkel, a judgement of the United States Supreme Court from 1906, Mr Gale submits there is no “jurisdiction” between himself and what he describes as the “Corporate Fiction trading as Tasmania Police”. Mr Gale supplied me with a copy of the case law relied upon. Leaving to one side the fact that it is not a precedent binding upon any Australian Court, to utilise its findings in the manner contended by the defendant requires a degree of intellectual contortionism which is untenable. The case involves a discussion of the application of the United States Constitution’s 5th Amendment to individuals as opposed to corporate entities. How it might be relied upon to establish a jurisdictional challenge based on a lack of privity of contract between a citizen of Australia and a state police service is unclear. Mr Gale’s oral submissions did not serve to elucidate his point. 

In any event, I note that as at 1 January 2004 the Police Service Act 2003 superseded the Police Regulation Act 1898. The Police Service Act 2003 created the Police Service in Tasmania as it is today. Tasmania Police is a statutory body and not a corporation as contended for by the defendant. Like many Government entities, the Tasmania Police Service has an Australian Business Number so that it might comply with its taxation and superannuation obligations. Such registrations do not render those bodies “corporate fictions” as submitted by the defendant. The ground set out at paragraph 8(vi) is without merit. ... 

Mr Gale contends as follows: “I challenge the jurisdiction. The trading entity Tasmania Police, trading under Department of Police Fire and Emergency Management (Tas), ABN 19173586474 does not have jurisdiction over myself, the sentient man, without their consent through lawful contract”. It seems the defendant has confused the powers of police and the jurisdiction of this Court. I have already disposed of the defendant’s arguments concerning the authority of police to perform their duties and the non-application of the law of contract. As for the jurisdiction of this Court, I again refer to and respectfully adopt the comments and findings of Magistrate McKee in Police v Nibbs (supra). I am satisfied no genuine jurisdictional issue is raised by this ground. 

(x) Self-Incrimination 

Mr Gale’s argument is summarised in the following passage: “Therefore, whether there are statutory laws compelling a person to submit to providing anything that may tend to incriminate him, the fact remains that a person has the legal right to not provide any material, whether verbal or tangible, if the production of that material would tend to incriminate that person”. 

The defendant appears to confuse the authority of police to lawfully undertake their duties, and the admissibility of certain evidence. The Evidence Act 2001 provides substantial protections in relation to the admissibility of evidence, particularly in criminal proceedings. It provides the Court with a broad discretion to exclude evidence that is highly prejudicial and not sufficiently probative, or for public policy reasons. Additionally, sections 128 and 128A of the Evidence Act 2001 specifically preserve the privilege against self-incrimination in appropriate circumstances. 

The arguments advanced by Mr Gale in this point are misconceived. 

(xi) Private Administrative Process and Estoppel 

As best I am able to understand the defendant’s contentions, Mr Gale submits: (a) He engaged certain members of Tasmania Police in a private administrative process; (b) That he served on said police officers documents and demands for information; and (c) That upon none of the police officers responding to the documents or demands for information he was entitled to proceed “in default” and consider all of the proceedings brought against him to be “settled in full”, “made in error”, “forgiven” or “not legitimate in the first place and ...therefore void or fully discharged”. 

Much of the documentation provided to the Court in support of the above argument is unintelligible. I observe that none of the defendant’s demands for information/documentation complied with the Right to Information Act 2009. Further, the defendant seems to have demanded that each of the charges brought against him be proved by police to his satisfaction and in accordance with his views concerning the law. 

Leaving to one side the fundamental difficulties associated with the defendant’s understanding of the law, he is not nor could he be involved in an “administrative” legal process. Administrative law regulates government decision making concerning individual matters. In each of the proceedings before the Court there is no administrative decision open to review. It is simply alleged that on various occasions Mr Gale has breached one or more of the laws of this State. These are not matters which can be concluded via default in the sense of a civil action. In the materials supplied in support of his submissions Mr Gale included a copy of the Bills of Exchange Act 1909 (C’wlth). While not specifically referred to in argument there is an earlier mention of the legislation with reference to “the private administrative process” at paragraph 8 of Mr Gale’s “affidavit”. There is no application of the Bills of Exchange Act 1909 to the unilateral communications apparently engaged in by the defendant and directed at certain police officers. The Bills of Exchange Act 1909 predominantly regulates the issue and use of promissory notes and can have no impact upon dealings between Tasmania Police and members of the public suspected of having committed an offence. 

(xii) Right to Subrogation 

This ground is based upon Mr Gale’s statement that he is the “beneficiary and general executor for the estate of his person”. If the defendant’s arguments were correct then he could transfer all of the rights, interests and liabilities in his person to another. It seems that, at least to the extent it furthers his ends, this is exactly what Mr Gale contends. By “subrogation” he purports to transfer his “liabilities” concerning the charges set out in the complaints to the Prosecutor. Again, I find that I can give no better reason for rejecting the defendant’s contentions other than to say they have no basis in law or common sense. 

(xiii) The Constitutional Point 

During oral submissions Mr Gale argued that while initially valid the Constitution of Australia was rendered a nullity when the Commonwealth of Australia became a privately owned US corporation. In support of this contention the defendant provided a significant quantity of documentation apparently obtained via a Freedom of Information request to the Department of the Treasury (Cth). 

A review of the materials provided by the defendant reveals the Commonwealth of Australia voluntarily sought registration with the Securities and Exchange Commission (US) in or about 2009 so as to provide information in support of the “Australian Government Guarantee Scheme”. The scheme being a mechanism whereby eligible deposit taking institutions (mainly banks) could apply to have unsecured debt guaranteed by the Commonwealth of Australia. Said scheme remained in place during a period of “heightened turbulence in international capital markets” – namely between November 2008 until March 2010. The scheme was obviously designed to increase confidence in Australia’s main financial institutions during said period of financial “turbulence”. 

In relation to this ground of argument I refer to the comments of Porter J in Nibbs v Devonport City Council [2015] TASSC 34 at paragraph 44: “A matter will not arise under the Constitution if it does not really and substantially, or genuinely, arise: see ACCC v CG Berbatis Holdings [1999] FCA 1151; (1999) 95 FCR 292 at 297; Danielsen v Onesteel Manufacturing Pty Ltd [2009] SASC 56 at [25]–[30]; Pham v Secretary, Department of Employment and Workplace Relations [2007] FCAFC 179 at [12]. A constitutional point must not be trivial or vexatious, or frivolous in the sense of being patently unarguable or completely devoid of merit”. cc No genuine constitutional point has been raised. The arguments proffered by the defendant are devoid of merit. Additionally, I note the operation of s 78B(1) of the Judiciary Act 1903. There is no evidence that notice has been served on the Attorney-General for the Commonwealth, or for the States of Australia. Thus, I cannot determine any Constitutional point of law, even if such had been properly raised. 

Conclusion 

I refer to and adopt the comments of Martin J, with whom the other members of the Court of Appeal agreed, in Kosteska v Magistrate Manthey & Anor [2013] QCA 105 at [17] where his honour said:

“This is not the first case in which Ms Kosteska’s argument has been advanced. It, and others like it, have wasted the time of the courts and opposing litigants, together with taxpayers’ money for some time. (This is not a peculiarly Australian problem. Similar fruitless cases have burdened the Canadian courts – so much so that Associate Chief Justice Rooke has examined in detail the characteristics, indicia and concepts of what he describes as Organised Pseudolegal Commercial Arguments.)”

And in NSW Bauskis v Wainhouse & Ors [2020] NSWCA 17 concerned Bauski' denial of jurisdiction regarding the seizure of a red cattle dog named “Duke” by officers of Fairfield Council from Bauskis’ residential property. 

Bauskis had made allegations of trespass to property and goods, conversion of a chattel (“Duke”) and sought the return of Duke together with damages. The parties sued were Fairfield Council, two of its officers who effected the seizure of “Duke”, and the State of New South Wales as the entity responsible for the police officers who attended the scene when Mr Bauskis refused the Council officers access to his property. 

The litigation begain when Bauskis refused the Council's dogs officer entry to his property after two dog attacks and stated he did not recognise the Council as a legal entity. Bauskis contacted the police, who attended the property. Bauskis reportedly said “[t]he Companion Animals Act is not the law and it does not apply to me”. 

 His submissions in this instance were:

(1) the case brought by the applicant in the District Court could only be determined by a jury; 

(2) there were procedural defects including the defendants not being required to “enter any pleas” such that the proceedings were to be described as a “Kangaroo Court”; 

(3) a contempt of court was committed by unidentified people in relation to the proceedings; 

(4) the District Court had no jurisdiction and could not proceed without a jury without the “clear and unequivocal consent of the parties to do so”. In this regard a report of the reasons of Forbes CJ in R v the Magistrates of Sydney [1824] NSWKR 3; [1824] NSWSupC 20 and the Supreme Court Procedure Act 1900 (NSW) was relied upon; 

(5) a submission under the heading “QUO WARRANTO” was advanced, relying upon, variously, the restoration of Charles II in May 1660, the Constitution of the Commonwealth, the maxim “if it ain’t broke, don’t fix it”, remarks attributed to Thomas Jefferson and Ecclesiastes 3:16.

Analytics

'What’s the Problem with Learning Analytics?' by Neil Selwyn in (2019) 6(3) Journal of Learning Analytics 11–19 comments

This article summarizes some emerging concerns as learning analytics become implemented throughout education. The article takes a sociotechnical perspective — positioning learning analytics as shaped by a range of social, cultural, political, and economic factors. In this manner, various concerns are outlined regarding the propensity of learning analytics to entrench and deepen the status quo, disempower and disenfranchise vulnerable groups, and further subjugate public education to the profit-led machinations of the burgeoning “data economy.” In light of these charges, the article briefly considers some possible areas of change. These include the design of analytics applications that are more open and accessible, that offer genuine control and oversight to users, and that better reflect students’ lived reality. The article also considers ways of rethinking the political economy of the learning analytics industry. Above all, learning analytics researchers need to begin talking more openly about the values and politics of data-driven analytics technologies as they are implemented along mass lines throughout school and university contexts.

'It's My Data! Tensions Among Stakeholders of a Learning Analytics Dashboard' by Kaiwen Sun, Abraham H. Mhaidli, Sonakshi Watel, Christopher A. Brooks, and Florian Schaub in Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (2019) 1-14 comments

Early warning dashboards in higher education analyze student data to enable early identification of underperforming students, allowing timely interventions by faculty and staff. To understand perceptions regarding the ethics and impact of such learning analytics applications, we conducted a multi-stakeholder analysis of an early-warning dashboard deployed at the University of Michigan through semi-structured interviews with the system's developers, academic advisors (the primary users), and students. We identify multiple tensions among and within the stakeholder groups, especially with regard to awareness, understanding, access and use of the system. Furthermore, ambiguity in data provenance and data quality result in differing levels of reliance and concerns about the system among academic advisors and students. While students see the system's benefits, they argue for more involvement, control, and informed consent regarding the use of student data. We discuss our findings' implications for the ethical design and deployment of learning analytics applications in higher education. Early warning dashboards in higher education analyze student data to enable early identification of underperforming students, allowing timely interventions by faculty and staff. To understand perceptions regarding the ethics and impact of such learning analytics applications, we conducted a multi-stakeholder analysis of an early-warning dashboard deployed at the University of Michigan through semi-structured interviews with the system's developers, academic advisors (the primary users), and students. We identify multiple tensions among and within the stakeholder groups, especially with regard to awareness, understanding, access, and use of the system. Furthermore, ambiguity in data provenance and data quality result in differing levels of reliance and concerns about the system among academic advisors and students. While students see the system's benefits, they argue for more involvement, control, and informed consent regarding the use of student data. We discuss our findings' implications for the ethical design and deployment of learning analytics applications in higher education.

'Big Data in Education. A Bibliometric Review' by José-Antonio Marín-Marín, Jesús López-Belmonte, Juan-Miguel Fernández-Campoy and José-María Romero-Rodríguez in (2019) 8(8) Social Sciences 223 comments

The handling of a large amount of data to analyze certain behaviors is reaching a great popularity in the decade 2010–2020. This phenomenon has been called Big Data. In the field of education, the analysis of this large amount of data, generated to a greater extent by students, has begun to be introduced in order to improve the teaching–learning process. In this paper, it was proposed as an objective to analyze the scientific production on Big Data in education in the databases Web of Science (WOS), Scopus, ERIC, and PsycINFO. A bibliometric study was carried out on a sample of 1491 scientific documents. Among the results, the increase in publications in 2017 and the configuration of certain journals, countries and authors as references in the subject matter stand out. Finally, potential explanations for the study findings and suggestions for future research are discussed.

'Ethical challenges of edtech, big data and personalized learning: twenty-first century student sorting and tracking' by Priscilla M. Regan and Jolene Jesse in (2019) 21(3) Ethics and Information Technology 167-179 comments

With the increase in the costs of providing education and concerns about financial responsibility, heightened consideration of accountability and results, elevated awareness of the range of teacher skills and student learning styles and needs, more focus is being placed on the promises offered by online software and educational technology. One of the most heavily marketed, exciting and controversial applications of edtech involves the varied educational programs to which different students are exposed based on how big data applications have evaluated their likely learning profiles. Characterized most often as ‘personalized learning,’ these programs raise a number of ethical concerns especially when used at the K-12 level. This paper analyzes the range of these ethical concerns arguing that characterizing them under the general rubric of ‘privacy’ oversimplifies the concerns and makes it too easy for advocates to dismiss or minimize them. Six distinct ethical concerns are identified: information privacy; anonymity; surveillance; autonomy; non-discrimination; and ownership of information. Particular attention is paid to whether personalized learning programs raise concerns similar to those raised about educational tracking in the 1950s. The paper closes with discussion of three themes that are important to consider in ethical and policy discussions. 

The last 10 years have witnessed an explosion of new educational technologies (edtech), some touting amazing potential to reach the next generation with new learning methods that will teach not only content, be it history, mathematics or engineering, but also intra- and inter-personal competencies, such as resilience and teamwork. The edtech sector is actively marketing these learning tools, especially to elementary and secondary schools, although the efficacy of technology enhanced learning is still under investigation. Edtech applications have appeared at a political, policy, and commercial moment favorable to the capabilities and advantages offered. The increase in the federal, state and local costs of providing K-12 education and government and voter concerns about financial responsibility generate interest in new techniques that promise to improve efficiency of educational operations. Focus on student achievement and the rankings of US schools with those of other countries has led to heightened consideration of accountability and results. Elevated awareness of the range of teacher skills, as well as variations in student learning styles and needs, has drawn attention to the value of understanding unique characteristics of students and teachers. As a result, the K-12 school environment is conducive to the promises offered by online software and edtech. Edtech companies recognize the huge market offered by K-12 education—an arena that has a vast and renewable population base, but also a particularly vulnerable population involving minor children who experience a range of developmental milestones during the K-12 years. 

This uptick in adoption of a variety of edtech applications at the K-12 level has also generated myriad policy debates, including proposed updates to existing federal laws and the introduction and adoption of numerous new state laws. Much of the policy debate is subsumed under the label of “privacy,” although there are a range of ethical issues associated with edtech applications that have not received the same amount of consideration as privacy, and some issues have been conflated with privacy. Privacy is certainly an issue, as the use of edtech entails collection of more, and more granular, information about students, teachers, and families, as well as administrative details regarding the functioning of educational institutions. Edtech applications enable sophisticated searching and analysis of collected information linking changes in the education arena to the larger debates about the challenges of big data generally. One of the most problematic aspects of edtech, and least addressed from a policy perspective however, involves the capability of edtech to deliver more personalized learning based on the needs and skill levels of individual students. 

Personalized learning applications are currently among the most heavily-marketed, exciting and controversial applications of edtech. These applications involve evaluating students likely learning profiles on applications that use big data to categorize individual learning styles and then direct appropriate learning activities to those students. Known under several labels—personalized learning, student-centered learning, and adaptive learning—they are advocated by edtech companies and foundations, including the Bill and Melinda Gates Foundation and the Chan Zuckerberg Foundation. In 2016, 97% of school districts surveyed by the Education Week Research Center indicated they were investing in some form of personalized learning (Herold 2017). Although exactly what types of programs constitute personalized learning is not always clear and whether and how much these programs incorporate edtech is hard to determine, RAND in the third of its reports on personalized learning cautions that the evidence for the effectiveness of personalized learning is currently weak and needs more research in a range of school settings (Pane et al. 2017). 

A critical ethical concern raised with personalized learning is whether such programs constitute tracking and sorting of students that might be considered discriminatory. The history of tracking in the United States is especially problematic, suggesting the need for caution when sorting children. Student tracking in the 1950s resulted in classrooms that were often divided by race, ethnicity, gender and class. Such tracking was glaringly obvious to parents, students, teachers and administrators—and thus the implications and wisdom of tracking became subjects of policy and social debate. In contrast, the student tracking that appears to be occurring in 2018 is hidden from the view of students, parents and even teachers as it takes place behind computer screens. The extent to which students might recognize they are being tracked through computer programs, and the impact that might have on learning outcomes is rarely discussed or researched. Similarly, the extent to which edtech software embeds subtle discrimination is also unclear, despite the current dialog about algorithmic bias. 

This article seeks to first analyze the range of ethical issues raised by the increased use of edtech and big data in school systems throughout the United States—how these issues are framed; whether the major concerns are receiving the appropriate level of attention and analysis; and what policy implications there are around how issues are being presented. Second, the paper briefly explores policy responses to big data educational innovations—what discourse has resulted; and what policy trends are emerging. Third, the paper is particularly interested in personalized learning systems and whether and how they might incorporate categories such as race, gender, ethnicity, and class, as well as their intersections, and whether discussions about these systems mirror the concerns of the policy and social debates in the 1950s about educational tracking. Finally, the paper closes with some themes that are important to consider in ethical and policy discussions addressing personalized learning systems.

Opacity

'Through a Glass, Darkly: Artificial Intelligence and the Problem of Opacity' by Simon Chesterman in (2020) American Journal of Comparative Law comments

 As computer programs become more complex, the ability of non-specialists to understand how a given output has been reached diminishes. Opaqueness may also be built into programs to protect proprietary interests. Both types of system are capable of being explained, either through recourse to experts or an order to produce information. Another class of system may be naturally opaque, however, using deep learning methods that are impossible to explain in a manner that humans can comprehend. An emerging literature describes these phenomena or specific problems to which they give rise, notably the potential for bias against specific groups. 

Drawing on examples from the United States, the European Union, and China, this article develops a novel typology of three discrete regulatory challenges posed by opacity. First, it may encourage — or fail to discourage — inferior decisions by removing the potential for oversight and accountability. Secondly, it may allow impermissible decisions, notably those that explicitly or implicitly rely on protected categories such as gender or race in making a determination. Thirdly, it may render illegitimate decisions in which the process by which an answer is reached is as important as the answer itself. The means of addressing some or all of these concerns is routinely said to be through transparency. Yet while proprietary opacity can be dealt with by court order and complex opacity through recourse to experts, naturally opaque systems may require novel forms of ‘explanation’ or an acceptance that some machine-made decisions cannot be explained — or, in the alternative, that some decisions should not be made by machine at all. 

Fintech Regulation and Disruption

'Dealing With Disruption: Emerging Approaches to FinTech Regulation' by Saule T Omarova in (2020) 61 Washington University Journal of Law and Policy 25 comments

 This article examines the emerging regulatory responses to an ongoing fintech disruption of traditional finance. Focusing primarily on the U.S. experience to date, it offers a three-part taxonomy of principal approaches to fintech taken by financial regulators: what I call the “experimentation” approach, the “incorporation” approach, and the “accommodation” approach. Within this framework, the article analyzes the pros and cons of establishing regulatory sandboxes and innovation hubs, issuing special fintech charters, and pursuing various regulatory-adjustment measures under a broad heading of RegTech.

'COVID-19 As a Force Majeure in Corporate Transactions' by Matthew Jennejohn, Julian Nyarko and Eric L. Talley comments

This paper surveys the use of pandemic-related provisions in Material Adverse Effects ("MAE") provisions in a large data set of publicly disclosed M&A transactions spanning the years 2003-2020. We document a trend towards greater use of such provisions, taking off particularly after the H1N1 crisis in 2009, and spiking again in late 2019 and early 2020. These terms are invariably located in the exclusions/carve-outs to the MAE, and they are overwhelmingly accompanied by "disproportionate effects" language that tends to dampen the effect of the carve out. There is little discernible statistical relationship between the inclusion of a pandemic-related carve-out and the inclusion of a reverse termination fee ("RTF") granting optionality to the buyer; but when an RTF is present, its magnitude tends to be smaller in the absence of any pandemic-specific carve-out, suggesting some degree of observational complementarity between these terms.

Palmer Inquiry

Andrew Palmer QC has been appointed to conduct a review of the Dangerous Goods Act 1985 (Vic) and associated regulations in response to concerns regarding chemical stockpiling after large chemical fires in West Footscray in August 2018 and Campbellfield in April 2019.

The media release refers to

a comprehensive review of the Victoria’s dangerous goods laws to help stamp out unsafe chemical stockpiling and keep Victorians safe.  ...   

Immediately following the 2018 fire, WorkSafe led a blitz on industrial premises to ensure dangerous chemicals were being stored correctly. Government agency investigations led to the discovery of waste chemicals stockpiled at 13 sites in Melbourne’s north last year. 

A WorkSafe-led taskforce has so far removed more than 13 million litres of the stockpiled waste chemicals, clearing four sites in Epping, three in Craigieburn and three in Campbellfield. The final three sites in Campbellfield are currently being cleared. 

The Labor Government last year introduced tough new penalties of up to 10 years in jail and fines of more than $6.4 million for rogue operators who recklessly manufacture, store, transport, transfer, sell or use dangerous goods in a way that places another person in danger of death. WorkSafe has charged Bradbury Industrial Services with 35 breaches of the Dangerous Goods Act in relation to five of the sites in Campbellfield and Craigieburn. ... 

 A final report and recommendations are expected to be delivered to the Government next year, with opportunities for public comment and stakeholder engagement later this year.

Blockchain

'The Law of Blockchain' by Georgios Dimitropoulos in (2020) Washington Law Review comments

Blockchain technology is a new general-purpose technology that poses significant challenges to the existing state of law, economy and society. Blockchain has one feature that makes it even more distinctive than other disruptive technologies: it is, by nature and design, global and transnational. Moreover, blockchain operates based on its own rules and principles that have a law-like quality. What we may call the lex cryptographia of blockchain has been designed based on a rational choice vision of human behavior. Blockchain adopts a framing derived from neoclassical economics, and instantiates it in a new machinery that implements rational choice paradigms using blockchain in a semi-automatic way, across all spheres of life, and without regard to borders. Accordingly, a global law and crypto-economics movement is now emerging owing to the spread of blockchain. 

This Article suggests that such a rational choice paradigm is an insufficient foundation for the future development of blockchain. It seeks to develop a new understanding of blockchain and its regulation through code according to the emerging ‘law and political economy’ framework. Blockchain should be understood as much more than a machine that enables the automation of transactions according to a rational choice framework. Blockchain should instead be understood as a technological infrastructure. Acknowledging the infrastructural dimension of blockchain technology may help identify a new role for the law in its interaction with blockchain, as well as for government in its interaction with the new technology. More precisely, identifying blockchain as an ‘infrastructural commons’ helps us recognize that law and regulation should not be relegated to the role of merely facilitating the operation of the invisible hand of the market by and within blockchain, but should rather acquire more active roles, such as safeguarding access on non-discriminatory terms to users, on a model with net neutrality and other public utility safeguards. The Article closes by proposing a ‘law and political economy’ framework for blockchain that is based on principles of publicness, trust, and interoperability.