Will AI create a K-shaped future for ADR?

When Signature Resolution CEO Dario Higuchi walks into conferences these days, one thing stands out:

"All of the booths are about AI," he said. "It's AI all over the place."

AI is rapidly spreading across the legal industry--from productivity tools to client management--and increasingly into dispute resolution itself.

"I think it's just kind of the reality of the legal landscape," Higuchi said. "I think it's up to us in ADR to start adapting or catching up. There's a lot of AI involved in every step of litigation."

The key question: Will humans stay in control?

JAMS CEO Kimberly Taylor says yes.

"Technology and AI are enhancing the dispute resolution process, but they are not replacing the role of experienced neutrals," she said. "The quality and subject-matter expertise of neutrals remain one of the primary drivers in selecting an ADR provider."

AI as decision-maker?

The debate sharpened last fall with the American Arbitration Association's rollout of AI Arbitrator, designed for lower-value, documents-only construction disputes. Trained on 1,500 real cases, the tool can issue outcomes and even recommend financial awards--if both parties agree to use it.

So far, adoption is minimal.

Judicate West Vice President Rosemarie Chiusano Aubert said that, at this point at least, AI as a decisionmaker is "more concern ... than usage."

"Big picture, I don't think AI will take the place of a third-party neutral," she said. "Perhaps on low-value disputes under $100K, parties may try it out to save money."

In February, AAA CEO Bridget McCormack said AI Arbitration "has officially one case on its docket."

Still, familiar concerns persist: hallucinations, bias and opaque decision-making. Even with party consent, courts may ultimately test enforceability.

The K-shaped split

A likely eventual outcome may be a K-shaped ADR market.

At the high end, complex and international arbitration is thriving. The International Chamber of Commerce reported near-record case volume at the end of 2025, with $300 billion at stake.

Aubert sees growing demand for customization:

Parties are "choosing by mutual agreement to deviate" from standard procedures in favor of "a willingness to tailor the process to achieve more efficient and cost-effective results."

That kind of nuance remains difficult to automate.

At the lower end, access is tightening. States--including California through SB 71--have raised small-claims limits as courts struggle to handle lower-value disputes. At the same time, access-to-justice debates are accelerating.

AI fills the gap--for now

Millions of people are already turning to AI tools for legal help, even as companies emphasize disclaimers. In November, OpenAI updated its policies to say it cannot provide legal advice and flags certain legal questions for attorney review.

Meanwhile, specialized tools are emerging to handle everyday legal tasks--drafting contracts, navigating divorce, resolving small-dollar disputes--echoing the role once filled by self-help books.

Full AI decision-making for consumers isn't here yet. But tools like AI Arbitrator hint at where things could go.

Efficiency vs. judgment

"The concept of streamlining the process for more straightforward documents-only disputes is a notable one for creating an efficient, lower-cost alternative," said ADR Services Inc. CEO Lucie Barron. "At the same time, there is a clear tension between the drive for efficiency and the need for thoughtful human analysis. Legal disputes often turn on nuance, including the context behind communications, industry practices, and credibility judgments that aren't always visible in the documents alone."

Her firm isn't offering a similar tool--for now--but is "always exploring opportunities to improve efficiency."

The privacy tradeoff

Higuchi is watching closely--but cautiously.

He worries about legality, confidentiality, and one unavoidable reality: AI systems need data.

That raises a potential tradeoff, especially for less-resourced parties--cost savings versus privacy.

"The arbitrations that are coming here are all confidential," he said. "There's a lot to be done with data, but there's a real question of confidentiality. Are the parties OK releasing some of their data?"