By Dr. Akshaya Kamalnath (The ANU College of Law)
The conversation about how artificial intelligence (AI) might affect various areas of law (and other areas of life) has, in recent months, centred around ChatGPT which is just one application of AI. In my recent article, ‘The future of corporate insolvency law: A review of technology and AI-powered changes’ [1] published in the International Insolvency Review, I take a broader view and assess how AI, and technology more broadly, has begun to transform, and will continue to transform corporate insolvency law. After examining recent instances of tech-enhanced insolvency processes in different jurisdictions, I argue that technology can improve efficiencies both prior to and during formal insolvency resolution processes. It, therefore, would be in the interests of every country to facilitate the adoption of technology at various points in the insolvency process.
Insolvency portals like Citius in Portugal and KOSTI in Finland have increased efficiencies by helping communications between parties and the court. In Colombia [2], Covid-19 catalysed technological enhancements to law and legal practices. This was particularly so in the context of insolvency law where the government approved the use of AI in the insolvency portal (MI) used by the bankruptcy court. While providing the basic features of e-platforms in Portugal and Finland, Colombia's AI powered system provides more sophisticated inputs into the process. Even the “basic features” like information sharing, document submission, and online access to information take on a very user-friendly format. Users can access all this on their smart phones with biometric authentication. When submissions are made through the system, the AI tool reviews them, verifies them against public databases, and makes a determination about whether the submission is complete. This determination then feeds into the court's decision about admitting a case into the formal insolvency process, thus speeding up the process. The system is even able to draft the decision which the judge can sign. These drafts are usually about requesting more information if the file is incomplete or a proposed decision. Thus, the AI tool seems to help with non-discretionary decision-making. At points where the AI made a mistake, for instance, in identifying foreign citizens, the human operators corrected this and the AI tool was able to learn from this. AI's ability to learn and improve over time shows that the AI tool may be useful at other points in the insolvency process. While valid concerns exist about using AI for discretionary and evaluative decisions because these sorts of decision-making cannot be coded into the system, the use of the AI tool in Colombia's bankruptcy court shows that there are efficiencies that can be gained by automating non-discretionary decisions.
Beyond formal court processes, mediation, which has been used in a lot of insolvency cases, is also an area where AI and other technologies can increase efficiencies. Indeed, technology has made inroads into ADR, and the use of technology in ADR has been termed online dispute resolution (ODR). ODR could refer to both technology-supported ADR and ADR that is performed by technology without human involvement. The former has already been used in other areas of law, particularly in the context of family law. One of the most well-developed examples comes from Canada—the British Columbia Civil Resolution Tribunal [3] which provides a diagnosis of the dispute, basic legal advice, and help with drafting in the form of customised templates. The advice in the initial phase goes beyond mere legal advice. Instead, it employs ADR concepts like the best alternative to a negotiated agreement for each party (BATNA). If this fails, then parties can choose to avail themselves of the services of a mediator. If ADR is not successful, then the Civil Resolution Tribunal will make a ruling. This system is currently available only for some categories of cases. One can imagine integrating such ODR systems within the type of e-portals for insolvency discussed earlier. That would allow for issues like creditor disputes which tend to hold up the insolvency process to have a cost-effective resolution. In smaller cases and in cases involving SMEs, the ODR system might even help to facilitate a settlement or plan for restructuring without court involvement (through a pre-insolvency procedure) or publicity. The latter point is significant in insolvency cases, where the stigma of formally entering an insolvency process often prevents early entry.
While technology has been able to greatly improve insolvency law and practice, and has the potential to further transform it positively, there are also risks involved. These risks should be identified and managed early on so that the benefits discussed immediately above are not lost. When AI is used in any context, the worry is that the decisions made by the AI are inaccurate or faulty, and biased. The former issue can be addressed by ensuring that AI is only used for minor decision-making at first and that parties have the right to appeal the decision. The issue of bias will need to be considered at the testing phase where the government, regulator, or court, as the case may be, is working with developers to develop/purchase an AI system to be officially used. Ultimately, the ability to appeal against an AI system's decision will again act as a check for this particular concern. Similarly, insolvency professionals, law firms, and others using AI in the market will need to test these tools before relying on them.
Privacy and data security is the other major concern that is often raised in the context of the use of AI. Even simply digitising court processes or other stages of the insolvency process gives rise to these issues. Further, legal obligations of different parties and those in charge of managing the electronic platform with respect to handling of personal data will need to be clarified. Further, the security and integrity of the systems (e-portals etc.) will need to be managed effectively. Further, requiring the use of technology in insolvency processes without proper training and education of court staff, regulators, and insolvency professionals will be counter-productive. It is important that the government in any jurisdiction seeking to formally introduce technology into insolvency processes must ensure that necessary training is provided.
Finally, we are also likely to see market innovation, especially from insolvency professional firms. Drawing from research in the context of the legal profession [4], I argue that insolvency professional firms are likely to consist of multi-disciplinary teams to accommodate such tech innovation to the practice.
*This article was originally published in the International Insolvency Review.
[1]https://onlinelibrary.wiley.com/doi/10.1002/iir.1512.
[2] https://www.ibanet.org/article/14AF564F-080C-4CA2-8DDB-7FA909E5C1F4.
[3] https://link.springer.com/article/10.1007/s10726-021-09734-1.
[4] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3688896.