Generative AI has changed what it means to be a self-represented litigant. AI tools allow individuals to quickly create large volumes of complaints, claims, communications, and court filings.
The legal system is currently facing an increase in the volume of AI-assisted applications. Some of them are well-founded, while others are not. This is particularly problematic in jurisdictions that deliberately encourage low-value litigation by individuals or unrepresented parties, such as employment dispute resolution forums (such as the Fair Work Commission) and administrative appeals courts.
Courts, tribunals, and businesses will need to adapt to accommodate the increasing number of disputes and proceedings arising from AI-assisted parties.
Japanese case: A warning example
In a recent US lawsuit, Nippon Life Insurance Company of America v OpenAI Foundationhighlights some of the challenges that arise when AI tools prompt self-proclaimed individuals to act legally inappropriately.
Nippon had previously settled a long-term disability lawsuit with a former employee, but the lawsuit was dismissed. Nippo’s complaint alleges that a former employee later pasted the attorney’s settlement email and court documents into ChatGPT. ChatGPT allegedly characterized the settlement as unfair, encouraged users to terminate their legal representation, created separate motions to reopen proceedings that had already been dismissed, and filed additional motions that were ultimately dismissed as frivolous.
Nippon is suing OpenAI for approximately US$300,000 in legal fees and US$10 million in punitive damages based on three principal causes of action under Illinois law:
- Tortious interference with the contract on the basis that ChatGPT induced or facilitated a breach of the settlement agreement by encouraging the pursuit of ultimately agreed claims.
- Abuse of process based on ChatGPT repeatedly provoking frivolous applications that impose costs and burdens on the other party. and
- Unauthorized legal action based on ChatGPT crossing the line from providing general information to providing customized and individualized legal assistance and drafts.
This case exemplifies the burden that AI-based applications can place on opposing parties. These include multiple frivolous applications, high legal costs, and protracted disputes that simple settlement agreements are designed to avoid. The liability of AI providers has not yet been tested in Australia, but the pattern of behavior described is not unique to the US.
Burden on courts and businesses
When unrepresented parties use AI to conduct legal proceedings, legal experts are present to contextualize and appropriately qualify the output arguments.
As a result, the system becomes taxed. As law enforcement agencies absorb an increasing number of claims, some of which are meritless, in addition to working to resolve legitimate disputes, companies must devote resources to addressing frivolous claims or claims that could have been more effectively resolved through standard grievance processes.
Resource pressures on decision makers and trading partners
Unrepresented parties can use AI to generate large amounts of material very quickly, and there may be no or limited oversight of the legal or factual accuracy of such material.
Businesses and decision makers must spend time and resources addressing such materials, regardless of whether they are accurate or contain merit. This is particularly acute for consumer-facing companies, where consumers may be more likely to litigate their complaints rather than try to resolve them through a company’s complaint resolution process.
The Chair of the Fair Work Commission has reported a significant increase in the amount of work associated with AI-assisted filing actions. This created a new problem. It is a problem that affects access to justice in other ways, by overwhelming the judiciary with unmerited materials and claims.
Risk to consumers
Consumers relying on outputs generated by AI will not be advised of all available options or the risks associated with each course of action. Consumers seeking redress regarding dispute risks:
- be affected by costs orders in connection with legal proceedings; or
- Failure to obtain a resolution that could have been resolved more efficiently and quickly through the grievance process.
Agency response: disclosure and case management
Courts and tribunals are beginning to respond through disclosure expectations and case management rules.
For example, the Federal Court of Australia recently issued a practice note stating that courts may require parties to disclose whether and how generative AI was used in litigation. The Fair Work Commission also recently published an exposure draft guidance note requiring parties to disclose whether AI was used in the preparation of submissions, with clear consequences for non-compliance (including potential adverse costs orders).
While these developments reflect a growing institutional awareness of this challenge, they do not address how companies, especially those who frequently appear in consumer-facing dispute resolution forums, should develop their own responses.
Practical steps for your business
There are practical steps you can take now to manage the impact of AI-generated unrepresented complainant materials.
Central to this is the grievance process. Identifying and resolving issues early, before they escalate into formal legal proceedings, could reduce the burden on companies to respond to AI-generated materials from unrepresented complainants.
In practice, this may include the following types of measures:
- Early triage of complaints: Empower frontline teams to quickly identify potentially escalated complaints, including those supported by AI-generated materials. These processes should facilitate rapid internal escalation and early resolution of high-risk complaints before they enter formal dispute channels.
- focus on important issues: Companies can consider complaints with the potential use of AI in mind and aim to focus their dispute resolution efforts on what they believe are the important issues being raised by the complainant. In some cases, it may be helpful to explicitly engage with complainants in this way. For example, refer to a 50-page complaint, note that the primary concerns appear to be X and Y, and explain your organization’s position on those issues.
- Addressing the root cause: Rather than getting into a confrontation with a particularly agitated complainer, companies should use customer feedback to quickly identify the root cause of the product or process that is causing the complaint and strive to address all customer pain points.
- Efficient response workflow: Develop processes and materials (including templates, workflows, etc.) to facilitate efficient, targeted and consistent responses to AI-assisted claims, particularly in large forums such as administrative courts and the Fair Work Commission.
