AI is not the lawyer, that maybe its greatest strength

JULITA MUSHATI AND FUNGAI CHIMWAMUROMBE

There is a particular kind of exhaustion that is familiar to the legal profession.
It lives somewhere between stacks of unread files, urgent client emails, court deadlines and the quiet pressure of knowing that one overlooked detail may alter the outcome of an entire matter.
The pressure to remain accurate, analytical and persuasive never disappears. Every word matters, every omission may carry consequences.
To many, legal work often appears polished and controlled, yet the profession has always demanded two things at the same time, speed and perfection.
Yet despite this, the legal profession is intellectually demanding, Lawyers absorb vast amounts of information, manage multiple files simultaneously and must maintain precision in the process. The workload is relentless even the most diligent lawyers can become mentally fatigued.
In an environment where time is limited but expectations remain exceptionally high, it is perhaps unsurprising that Artificial Intelligence (“AI”) has increasingly found its way into modern legal practice.
It is within this reality that AI has emerged as both a promising tool and a growing ethical concern within modern legal practice. Around the world, lawyers are increasingly using AI-powered tools to summarize documents, simplify complex language, assist with research, organize information and improve efficiency in drafting. In many respects, these technologies are not inherently harmful to the profession. Used responsibly, AI may assist legal practitioners in saving time, reducing administrative burdens and making legal services more accessible.
However, the growing use of AI within legal practice raises an important legal question: to what extent may lawyers rely on AI-generated material before it becomes professionally dangerous? While AI may improve efficiency, the legal profession remains built upon accuracy, originality and personal accountability. Courts do not assess whether a document was produced quickly. They assess whether it is truthful, legally sound and supported by genuine authority.
The danger of relying on unverified AI- generated material was recently highlighted in the Supreme Court case of Pulserate Investments (Pvt) Ltd v Andrew Zuze and Others [SC202/25], the court rejected heads of arguments submitted, declaring them a nullity as they contained fictitious cases generated by AI, consequently, the court ordered the appellant to bear the costs. The judgment sent a clear warning that unverified AI-generated research has no place in judicial proceedings.
This precedent underscores a strict judicial stance that, the reliability of legal submissions depends entirely on verifiable, existing authority, not on the perceived credibility of automated outputs.
Consequently, the risks associated with AI extend beyond inaccurate research. AI platforms are capable of producing convincing but entirely fabricated legal authorities, commonly referred to as “hallucinations.”
If such material is placed before the court without verification, legal practitioners risk misleading the court, damaging their professional credibility and exposing clients to prejudice. In serious circumstances, this may amount to professional negligence or even contempt of court where false information is deliberately or recklessly presented as genuine legal authority as the foundation of law rests in the use of legal authentic precedents.
Perhaps the healthiest way to understand AI within legal practice is to think of it as a navigation device, where it assists legal practitioners in organizing information, simplifying complexity and improving efficiency, but it should never replace legal reasoning itself or authentic authority. General AI models are not specialized legal tools and often lack access to verified legal repositories, resulting in hallucinated or non-existent authorities. The issue is therefore not the use of AI, but its substitution for independent legal reasoning and professional judgment.
At present, there is no law in Zimbabwe expressly prohibiting the use of AI in legal practice. However, the Legal Practitioners (Code of Conduct) By-Laws, 37 of 2018, particularly section 3 subsections (20) to (22), reinforce the duty to uphold integrity, competence and the good standing of the profession.
Given that legal practitioners are entrusted with clients’ livelihoods, assets and rights, a higher standard of accountability is required, and lawyers remain fully responsible for the accuracy of all court submissions, regardless of technological assistance.
This raises an important question: while technology should be embraced to improve efficiency in legal practice, should Zimbabwe consider regulatory frameworks to govern the use of AI, particularly mechanisms to verify legal authority and ensure reliability in AI-assisted research?

Mushati is a legal intern at Zenas Legal Practice and can be contacted for feedback on julita@zenaslegalpractice.com and whatsapp 0772306088
Chimwamurombe is a registered legal practitioner and Senior Partner at Zenas Legal Practice and can be contacted for feedback at fungai@ zenaslegalpractice.com and WhatsApp 0772 997 889.