AI in the Courtroom

AI in the Courtroom

Draft legal submissions
Summarise case law
Predict outcomes based on past judgments
Transcribe hearings (excellent use)
Flag inconsistencies in evidence (questionable use)
Some judges and lawyers use tools like Copilot Chat, a secure AI assistant provided by Microsoft, to help with research and writing (help is fine, reliance is not)
In South Africa, lawyers submitted fake case law generated by AI
In England, the High Court warned that AI misuse could damage public trust
AI tools make stuff up: citing facts or cases that don’t exist.
Lawyers using such tools really must verify all the facts and citations – so much so, that they may as well do their own research the old fashioned way in the first place!
Lawyers cannot blame the AI if something goes wrong
I cannot imagine a lawyer (none I know anyway) doing so, but uploading confidential client date to public AI platforms in order to get case law and citations will be a breach of privacy rules.
Judicial Guidance (2025) outlines how judges should handle AI generated content (it’s likely to be a breach of any lawyer’s code of conduct if they use AI submissions that are false in court)
Courts are encouraged to scrutinise unfamiliar citations and persuasive but flawed arguments
Ethical standards are being updated to reflect AI risks
There is also talk of creating independent bodies to assess AI tooks before they’re used in legal practice (overkill, a lawyer should be able to determine their own use of such tools without yet another independent body interferring).
Treat it like any other tool: understand its limits and verify its output
Don’t rely on it blindly: AI can’t replace legal judgment or ethical responsibility
Stay informed: read the latest guidance and case law on AI use
Protect your own and clients’ data: use secure platforms and avoid uploading sensitive information to public tools.
AI in the Courtroom

Leave a Reply

Scroll to top