Traditional code-plagiarism detection tools relied on a token-level view of source code to detect "similar" token subsequences among soource files. More sophisticated tools attempt to compare program structure to identify common structural elements. The object of this project is to evaluate the effectiveness (or otherwise) of structure-based plagiarism-analysis techniques to detecting code produced by generative-AI tools such as ChatGPT and develop a tool based on these techniques that maximises the detection of generated code.