Trust, not speed, will define legal AI in UAE: Expert

Artificial intelligence (AI) is rapidly reshaping legal practice across industries in the UAE and wider GCC. But unlike many markets, legal systems in this region are genuinely distinct, jurisdiction by jurisdiction — a nuance that raises the stakes for how legal AI is built.

The focus must remain on building trust alongside innovation. Trust in legal AI rests on three vital pillars including privacy, provenance and accountability.

This was stated by Lara Salem, Head of Content & Data Strategy LexisNexis Middle East in an exclusive interview with Gulf Today in Dubai.

“As governments modernise their legal systems and accelerate digital transformation, AI adoption is no longer the central question. The conversation has shifted to something far more important: can legal professionals trust the AI they rely on?,” she said while sharing insight.

She pointed that for legal professionals, trust matters more than speed.

Lara Salem, Head of Content & Data Strategy LexisNexis Middle East.

“Every AI-generated response has the potential to influence legal advice, commercial decisions and regulatory compliance, making accuracy, transparency and accountability essential. Realising the full value of legal AI requires solutions grounded in authoritative regional legal intelligence, supported by robust governance and meaningful human oversight. When these foundations are in place, AI becomes more than a productivity tool — it becomes a trusted partner that enables better-informed decisions, greater efficiency and a more resilient legal ecosystem.” The first pillar, Lara Salem said, is privacy: encryption, security and confidentiality by design. It is necessary, but it is no longer enough to set a provider apart — every serious AI vendor now claims the same thing.”

“The second, rarely discussed with the same seriousness, is provenance. Where did this answer come from? Is it grounded in current legislation and case law, or in whatever the model absorbed during training? A confident wrong answer is more dangerous than an honest “I don’t know” and general-purpose AI has no mechanism to tell the difference. Purpose-built legal AI does. At LexisNexis, our Protégé platform is built on agentic AI paired with authoritative, jurisdiction-specific legal content, so every output can be traced back to its source.

“The third is accountability, and this is where the industry gets uncomfortable. Governance frameworks matter, but they can become a way to outsource judgement to a policy document – and judgment can’t be outsourced. If an AI-assisted brief is wrong, the lawyer who filed it is accountable, not the software, not the governance committee.”