October 16, 2026, (Inside AI) — Pakistan's government has confirmed it is reviewing whether its criminal statutes are adequate to prosecute crimes planned or carried out with artificial intelligence, after two FBI-triggered investigations led to arrests in Punjab province.
Law Minister Azam Nazeer Tarar said existing cybercrime provisions and general criminal law cover some AI-assisted offenses, but new legislation may be needed as the technology evolves. His comments followed confirmation that the National Cyber Crime Investigation Agency (NCCIA) opened two cases after alerts from the US Federal Bureau of Investigation.
In the first case, a 17-year-old student was arrested in Bahawalpur over an alleged plot to poison his father. In the second, investigators are examining a separate teenager accused of planning an attack against someone who had insulted his father. Both allegations remain subject to the legal process, and no convictions have been recorded.
The cases matter beyond Pakistan's borders. They test whether a country's criminal code can handle conduct that is planned with a chatbot rather than a conspirator, and they expose a widening gap between how fast AI tools spread and how slowly courts and police adapt. Pakistan is not alone in confronting this question, but its reliance on FBI tips shows how cross-border intelligence sharing is now shaping domestic AI enforcement.
Read: India Supreme Court Proposes Ban on AI in Judicial Decision-Making
Tarar framed the review as a routine legal audit. He said legislation must evolve alongside technological developments and that relevant provisions could be introduced if gaps were identified. He did not commit to a timeline or a draft bill.
Do Existing Laws Already Cover AI Crimes?
Digital rights advocates are not convinced that Pakistan needs a dedicated AI statute. Usama Khilji, director of the digital rights organization Bolo Bhi, argued that conspiracy and attempted murder provisions already cover many of the offenses under investigation. He instead pointed to the responsibilities of AI companies, calling for stronger safeguards on how their systems handle dangerous requests.
His position reflects a broader debate in technology law. Countries including the United Kingdom and Canada have so far chosen to amend existing statutes rather than write standalone AI criminal codes. The European Union took a different route with its AI Act, though that regulation focuses on product safety and market rules, not criminal liability.
High Court advocate Minahil Ali took the opposite view. She said Pakistan needs clearer and more enforceable safeguards against emerging risks from advanced AI systems. She called for updates to criminal law, specialized cyber-forensics training for investigators, and stronger procedures for establishing criminal intent.
That last point is the hardest technical problem. A chat log showing a user asking an AI model about poisons does not by itself prove intent. Prosecutors would need corroborating evidence such as purchases, communications, or physical preparations. Experts warned that investigations should not rest on isolated AI chat records or automated safety alerts alone.
The NCCIA's handling of the two Punjab cases will serve as an early test of that standard. If charges proceed on chat logs alone, defense lawyers are likely to challenge them. If investigators build cases on physical evidence, the prosecutions could set a template for how South Asian courts treat AI-assisted planning.
Pakistan's review also intersects with a regional trend. India has issued advisories on deepfake-related crimes, and Bangladesh has expanded its cyber tribunal's remit. None of these countries has passed a comprehensive AI criminal law, leaving enforcement to patchwork rules.
For now, Tarar's ministry is assessing how existing laws apply and whether additional legislation is required. The outcome will depend on whether the government treats AI as a new category of crime or simply as a new tool used to commit old ones.
