Artificial Intelligence has quietly entered legal practice in Pakistan. It did not arrive through any law or policy. It came through convenience.
Today, many lawyers are using tools like ChatGPT, Gemini and Claude etc to draft pleadings, summarize judgments, and refine legal arguments. These tools are fast, helpful, and increasingly common. But while their use is growing, something important is missing.
There are no rules.
A Silent Shift in Legal Practice
The legal profession has always adapted to technology. First came typewriters, then computers, and later digital legal databases. Each step improved efficiency without changing the core responsibility of lawyers.
Artificial Intelligence is different. It does not just assist in finding information. It generates new content.
That content can look polished and convincing. It can also be completely wrong.
This is where the risk begins.
Risks of AI in Court Proceedings
The global legal community has already seen the dangers of careless AI use.
In the well-known Mata v. Avianca, Inc. case, lawyers submitted court documents containing fake legal citations generated by AI. The citations looked real, but they did not exist. The court imposed penalties, and the incident became a warning for lawyers everywhere.
This was not a technical error. It was a failure of verification.
In Pakistan, where courts rely heavily on written submissions and cited precedents, such an error could easily mislead proceedings. Even one incorrect citation can waste time or affect a decision.
A System Without Safeguards
Despite these risks, Pakistan has no formal framework to guide the use of AI in legal practice.
- The Pakistan Bar Council has not issued any guidelines.
- The Supreme Court of Pakistan has not introduced any practice directions
- High Courts have also remained silent.
This creates a situation where every lawyer is left to decide on their own how to use AI tools. Some use them carefully. Others may rely on them too much. There is no shared standard.
This silence is not harmless. It is a gap.
How Other Countries Are Responding
Other jurisdictions have started addressing this issue in a practical way.
In the United States, the American Bar Association has clarified that lawyers can use AI, but they must verify all content and remain fully responsible for their work. [Reference]
In the United Kingdom, the Bar Council of England and Wales has warned lawyers about the risks of false outputs and stressed the need for human oversight. [Reference]
India presents a closer comparison. The Bar Council of India has not yet issued national rules, but several courts have taken a cautious approach. Some have restricted AI use in judicial work, while others have emphasized strict verification.
Across all these systems, one idea is clear: AI can assist, but it cannot be trusted blindly.
Need for AI Regulation in Pakistan
Courts are not ordinary workplaces. They decide rights, property, and personal freedoms.
If AI begins to shape legal arguments without safeguards, it can affect the fairness of proceedings. The risk is not just technical. It is institutional.
Pakistan has often responded to new technologies after problems arise. In this case, waiting may not be wise.
AI is already being used. The system is already changing.
A Practical Way Forward
Pakistan does not need strict laws at this stage. It needs clarity.
Basic guidelines can make a big difference. Lawyers should verify all AI-generated content before submitting it. They should remain fully responsible for everything placed before the court. In some cases, it may also be useful to disclose when AI has played a major role in drafting.
Training is equally important. Lawyers need to understand both the benefits and the limits of these tools.
These steps are simple, but they create accountability.
The Case for a LegalTech Committee
A strong starting point would be the formation of a specialized committee within bar councils.
This committee should include AI Experts, lawyers with IT backgrounds, professionals working in LegalTech and LegalAI, and legal academics familiar with technology law.
Its role would be to study how AI is being used, identify risks, and suggest practical guidelines. It can also bridge the gap between the legal profession and fast-changing technology.
This approach allows regulation to grow with experience instead of reacting to mistakes.
The Real Question
Artificial Intelligence will continue to spread in legal practice. That is certain.
The real issue is whether the system will guide its use or ignore it. For now, the answer remains uncomfortable.
No one is regulating AI in Pakistani courts.
And that is precisely why the conversation must begin now.
