Alan Ma warns that hallucinations are just the tip of the iceberg when it comes to AI concerns
- Charts the next phase of AI verification, highlighting that information bias, prioritisation logic and omitted information and argument are harder to detect than hallucinations.
- Concludes the legal profession can adapt to overcome these hurdles.
The legal profession is navigating an essential period of professional recalibration. The tools have changed, and the speed, scale and confidence with which legal information can now be generated have changed with them. Yet the fundamental requirement remains the same: independent human verification is still central to legal integrity. The profession’s first serious encounter with generative artificial intelligence (AI) was dominated, understandably, by hallucination. Fabricated cases, invented quotations and non-existent legal propositions strike directly at the heart of professional credibility. Once courts began openly confronting fake authorities, the profession’s response was immediate: verify, supervise and take responsibility.
Yet the discussion has already started to evolve. The issue is no longer confined to whether AI systems can
