- Research in the US shows that the number of self-represented civil lawsuits rose from 11% (2005–2023) to 17% after AI became popular, accounting for most of the increase in total lawsuits.
- The rate of lawsuits showing signs of being AI-generated rose from nearly 0% in 2022 to 18% in 2026.
- AI-assisted lawsuits do not show lower quality: there is no increase in dismissal rates or processing times.
- However, the volume of documents and actions in each case has increased sharply, putting growing pressure on the courts.
- AI helps expand legal access for those who previously could not afford a lawyer or did not know they had the right to sue.
- Self-represented cases are primarily simple disputes such as employment discrimination or foreclosures.
- Lawyers’ workloads have not decreased; the number of cases with lawyers remains stable or has increased slightly.
- This suggests that AI expands the “legal market” instead of replacing lawyers.
- However, in the UK, lawyers have noted a “wave” of AI lawsuits with long, complex, but low-quality content.
- UK employment tribunals are severely backlogged, with some cases having to wait until 2028.
- AI causes litigants to have excessively high expectations of winning, making the mediation process difficult.
- Outdated legal systems struggle to meet the sudden surge in AI-generated records.
- AI can increase costs for businesses due to having to process larger volumes of documents.
- The US–UK difference may come from stricter case-filtering standards in US federal courts.
- General trend: there is huge latent legal demand and AI is activating it.
📌 AI both expands access to justice and creates systemic pressure. In the US, the rate of self-represented lawsuits increased to 17% and 18% of filings show AI signs but without a drop in quality. Conversely, the UK faces backlogs until 2028 due to a “flood” of AI records. Conclusion: AI does not replace lawyers but expands legal demand; however, if infrastructure is not upgraded, the benefits of access to justice could be negated by system overload.
