Review launched after 'concerning' AI parole error

An investigation into the potential use of artificial intelligence in parole decisions is being conducted following an alarming error in the case of Susan Neill-Fraser, a high-profile murderer. Tasmania's justice department announced the review, which follows Neill-Fraser's release from parole in 2022 after serving 13 years for killing her partner Bob Chappell in 2009.

The parole board had imposed conditions restricting Neill-Fraser from speaking to the media or claiming innocence, which were later deemed invalid and without proper procedural fairness by the Supreme Court of Tasmania.

The review aims to assess the degree to which AI may have influenced past parole board decisions in Tasmania. Justice department spokesperson confirmed that the department cannot comment further while the review is ongoing. The parole board operates independently from the department.

Greg Barns, chair of Prisoners Legal Service Tasmania, expressed concern over the conditions placed on Neill-Fraser and the fact that the case law cited in her case was not verified. His organization is conducting its own audit of parole board decisions over a two-year period to check for any signs of AI use or errors. Sarah Schwartz, from the Human Rights Law Centre, which represented Neill-Fraser, called the use of AI in generating the parole conditions deeply troubling, stating that it severely restricted her client's fundamental freedoms.

The parole board declined to comment on individual matters and could not provide details on the extent of its AI usage. Neill-Fraser, who has failed in multiple appeals against her conviction, intends to continue fighting for her exoneration. Attorney-General Guy Barnett has written to the parole board seeking an explanation, emphasizing that the mistake is unacceptable.