top of page
Search

How AI is Changing the Legal Profession

Abstract

Artificial Intelligence (AI) is rapidly transforming the legal profession, particularly in the area of legal research and document review. This article examines how legal professionals, from candidate attorneys to veteran practitioners, perceive AI adoption in terms of efficiency, ethics, and morality. It further explores how these perceptions shape concerns about job security and professional identity. Findings suggest that while AI offers substantial opportunities for efficiency and upskilling, it also raises ethical dilemmas and anxiety about the future of legal work, particularly for those in early career stages.


Introduction

Picture this: a client steps into a law office, expecting wisdom distilled from years of study and experience. Instead, the first draft of their contract was produced by an algorithm in seconds. Would you see that as a triumph of technology, or a compromise of trust?


The integration of Artificial Intelligence into legal practice is no longer a distant possibility, it is an emerging reality. Globally, tools such as Harvey, Casetext, and Lexis+ AI have begun reshaping the way lawyers conduct research, draft documents, and manage cases (Wolters Kluwer, 2024). In South Africa, while adoption lags behind international trends, firms are beginning to experiment with AI solutions to meet the demands of increasingly competitive markets.


This article explores three interconnected dimensions of AI adoption in the legal profession: efficiency, ethics and morality, and job security. It aims to capture the nuanced perspectives of both veteran attorneys and candidate attorneys, offering insights into how these dynamics might influence the future of legal practice.


Efficiency: Evident Gains, Uneven Maturity

Empirical snapshots point to real productivity benefits. According to the American Bar Association’s (2024) Legal Technology Survey Report, AI adoption in legal firms rose from about 11% in 2023 to 30% in 2024, with larger firms (100+ lawyers) showing usage at around 46%.


Thomson Reuters (2024) found that professionals broadly expect AI to apply to legal work (≈81–85%) and estimate significant time savings on routine tasks. Similarly, Wolters Kluwer’s (2024) Future Ready Lawyer Report highlighted that weekly use of Generative AI among corporate legal departments and law firms is becoming commonplace.


Yet implementation remains early-stage and uneven. A “human-in-the-loop” model is still essential to curb hallucinations and context errors (LawNext, 2024).


Ethics and Morality: Reliability, Bias, and Accountability

AI raises persistent ethical questions:

  • Confidentiality & Privilege. The Solicitors Regulation Authority (2023) emphasises the duty to protect client data when using AI, principles that align with South African expectations even without codified standards.

  • Reliability & Hallucinations. Studies confirm that hallucination risk persists even with retrieval-augmented systems, underscoring the need for verification (De Rebus, 2024).

  • Bias & Fairness. AI trained on historic legal text can embed inequities, raising moral concerns.

South African courts have already condemned fabricated AI-generated citations, signalling zero tolerance for unverified outputs in pleadings (South African Judiciary, 2024).


Job Security and Professional Identity

Candidate attorneys face disruption as AI automates research and drafting tasks once critical for training. Many fear fewer entry-level opportunities and weaker skill development (ABA, 2024).


Veteran attorneys, while less concerned about outright job loss, confront re-skilling pressures. They must integrate AI into workflows and advise clients on AI risks. Sector reports forecast shifts in billing and staffing models, even as overall demand remains strong (Thomson Reuters, 2024).


The South African Context

As of May 2025, the Law Society of South Africa’s Annual Report confirmed that no formal AI guidance had yet been adopted due to internal committee differences (LSSA, 2025). Nevertheless, practitioner journals such as De Rebus (2024) and recent judicial commentary highlight a de facto expectation: verify outputs, preserve confidentiality, and retain human accountability.


Bridging the Veteran to Candidate Divide

The literature suggests three steps to balance efficiency with ethical training:

  1. AI Literacy and Playbooks. Clear internal guidelines on when and how to use AI (LawNext, 2024).

  2. Preserve Foundational Training. Ensure candidate attorneys still practise manual analysis alongside AI (Wolters Kluwer, 2024).

  3. Ethics-by-Design. Build confidentiality safeguards and require human sign-off (SRA, 2023).


Recommendations

  • Firms: Adopt “human-in-the-loop” standards, rebalance training, and explore new pricing models.

  • Regulators: Finalise South African ethical guidance aligned with global standards.

  • Practitioners: Develop AI literacy, data governance fluency, and client-facing advisory skills.


Conclusion

AI is neither a threat to eradicate nor a cure-all for inefficiencies. For candidate attorneys, it disrupts apprenticeship but also creates new opportunities for tech-enabled practice. For veteran attorneys, it is a prompt to lead ethically and strategically. In South Africa, the future will depend on how regulators, firms, and professionals balance efficiency gains with ethical obligations and professional accountability.


References


This article is authored by Megan Lombaard, Law Student from South Africa and Trainee of Lets Learn Law Legal Research Training Programme. The views and opinions expressed in this piece are solely those of the author.


 
 
 

Comments

bottom of page