Legal AI Trust Rises With Use, AAA Survey of 557 Professionals Shows

Legal professionals who regularly use AI tools express significantly higher trust in the technology than colleagues who avoid it, according to new research examining how hands-on experience reshapes attitudes toward artificial intelligence in legal practice.

A study of 557 U.S. arbitration professionals released by the American Arbitration Association and Jus Mundi found that daily AI users report a trust score of 3.42 out of 5, compared with just 0.83 among non-users. The research, titled The State of AI in U.S. Arbitration 2026, was first reported by the organizations on September 23.

Why it matters

The findings suggest that firsthand experience—not abstract concerns—drives confidence in legal AI. As law firms debate adoption strategies, the data indicates that controlled exposure and training may be more effective than waiting for the technology to mature. The generational divide in usage also signals a fundamental shift in how legal work will be structured as junior lawyers who grew up with AI advance in their careers.

Experience Changes Risk Perception

While trust increases with use, the nature of practitioners' concerns evolves rather than disappears. Non-users focus predominantly on ethical issues, while regular users concentrate on practical challenges like confidentiality and protecting sensitive client information.

Accuracy remains the dominant concern across all experience levels, with 78% of respondents citing inaccuracies and hallucinations as a primary worry. This consistency suggests that even experienced users recognize fundamental limitations in current AI systems.

Generational Divide in Adoption

Career stage emerged as the strongest predictor of AI use. More than half of respondents with fewer than three years of legal experience use AI daily or weekly, and fewer than 5% rule out future adoption. Among practitioners with 15 or more years of experience, regular usage drops below 30%, and nearly 24% say they have no plans to adopt the technology.

Reshaping Legal Work

Respondents anticipate AI absorbing labor-intensive tasks including document review (51%), proofreading and cite-checking (47%), timeline creation (43%), and legal research (40%). Meanwhile, 52% expect strategic judgment to become more valuable as routine work shifts to automated systems.

This transition could compress the traditional learning curve for junior lawyers, requiring higher-level skills earlier in their careers. Despite this shift, half of respondents reported little to no concern about AI replacing aspects of their role, while only 13% expressed high concern.

What Practitioners Want

To accelerate responsible adoption, respondents identified institutional guidelines and best practices (30%) and arbitration-specific AI tools (24%) as top priorities. Cost reduction and demonstrated return on investment ranked near the bottom, suggesting practitioners view guidance and appropriate tooling as more pressing barriers than economics.

Firm size showed limited correlation with realized value. Sixty-six percent of boutique-firm respondents reported significant time savings, compared with 50% at global firms and 51% at mid-size firms, indicating that organizational agility may matter more than resources.

"The more legal professionals work with AI tools, the better they understand their capabilities, where they can deliver value, and what is needed to use them effectively and responsibly," said Linda Beyea, AAA group vice president.

The complete study was released by the American Arbitration Association and Jus Mundi.