August 28, 2026 | 11:31 pm
TEMPO.CO, Jakarta - The adoption of artificial intelligence (AI) for judiciary practices sparked backlash from digital practitioners, citing a sharp dissonance between algorithms and the principles of justice. In several countries, including the United States, AI is not only used to analyze case files, but to construct arguments, predict risks, and support decision-making.
The Chair of the Communication and Information System Security Research Center (CISSReC), Pratama Persadha, asserted that AI does not function like a human judge who arrives at a decision through a series of traceable legal considerations.
“AI models basically process statistical patterns from data and generate output based on probabilistic relationships learned during their training,” he told Tempo on Wednesday, August 26, 2026.
In the United States, for example, risk assessment algorithms like COMPAS, short for Correctional Offender Management Profiling for Alternative Sanctions, have been used to help assess the likelihood of someone committing a crime again. The analysis of over 7,000 COMPAS scores in Broward County, Florida, at that time indicated accuracy issues and error gaps among racial groups. In that analysis, the prediction of recurring violence was only correct about 20 percent of the time.
In Colombia, the use of ChatGPT by judges that surfaced in early 2023 sparked discourses. A judge in Cartagena admitted that the OpenAI-owned chatbot was used to generate questions in a case involving a child with autism regarding medical care financing.
Pratama also cited another example, the lawsuit filed by US citizen Roberto Mata against Avianca airlines in the United States in 2022. Using ChatGPT, Mata's lawyer unintentionally included six legal precedents that had never existed. This mistake led the lawyer and the law firm to be fined $5,000. “In a legal context, human presence as the ultimate decision-makers is crucial,” Pratama said.
Black Box AI Versus Legal Integrity
The risk of integrating smart tech into the legal sector is inseparable from the "black box" nature of AI: they can only provide specified answers that may not necessarily be accompanied by correct—let alone justifiable—reasonings.
This responsibility must not be placed solely on the algorithm. If, ultimately, one must adopt AI use for a legal proceeding, the system must not only be accurate, but the data must be traceable. The model must also be testable for bias, errors, and limitations of its application.
“The system must be auditable, documented, monitored, and placed within a human oversight mechanism,” he said.
Echoing the sentiment, the Executive Director of Information and Communication Technology (ICT), Heru Sutadi, said all parties must ensure the identity of AI users in the legal domain. “To ensure accountability for those who use the outputs of AI,” he said. “AI recommendations must also not be the sole basis for decisions concerning someone's rights.
The issue of AI adoption in justice was addressed in one of the sessions at the Kaspersky Asia Pacific Cyber Security Weekend 2026 in Guangzhou, China, in early August 2026. Legal practitioners as well as corporate lawyers, Janson Tan, from Tan & Tan Law Chambers, mentioned that legal decisions require evidence, transparency, accountability, and reasons.
As also discussed in Tempo's premium report, 'Jika Putusan Hakim Memakai Akal Imitasi' (When Court Rulings Use Artificial Intelligence), the aforementioned requirements prove to be difficult for AI, which tends to prioritize speed, automation, optimization, and the ability to work on a large scale.
“When the two meet, what then arises is an accountability gap,” Tan said.
Read: How Indonesia Pushes AI-Backed Surveillance on Foreigners
Click here to get the latest news updates from Tempo on Google News
Tech Giants Urge Global Response to AI Cybersecurity Threats
11 jam lalu
Concerns are growing over global cybersecurity after AI models escaped test environments and breached multiple organizations.
How Indonesia Pushes AI-Backed Surveillance on Foreigners
23 jam lalu
Indonesia is promoting the use of digital technology, including artificial intelligence (AI), to strengthen surveillance of foreign nationals.
Fact Check: Does the mRNA Dengue Vaccine Contain AI or Microchips?
2 hari lalu
Fact check - mRNA-based dengue vaccine candidate doesn't contain chip & doesn't alter DNA. AI is used in research process, but not included in vaccine
Germans Swap Friends for Screens in Their Free Time
2 hari lalu
Germans are spending less of their free time socializing with friends and more on individually planned and digital activities.
Australia's Music Chart Bars Fully AI-Generated Music
3 hari lalu
The ARIA chart distinguishes between artists who use AI as a creative tool and songs largely generated by machines.
Hyundai, KIA Workers Stage First Full-Scale Strike in South Korea in a Decade
5 hari lalu
Hyundai & KIA workers in South Korea stage their first full-scale strike in a decade, demanding higher wages, job security and a higher retirement age
Fact Check: Photos Claim to Show Israeli Soldiers Captured by Iran
5 hari lalu
Fact check - These photos of Israeli Soldiers captured by Iran are all AI-generated.
Fact Check: Video of US Troops Invading Iran Is AI-Generated
6 hari lalu
Tempo fact check team verified a viral video claiming to show United States troops landing on Iranian territory and found it to be AI-generated.
Apple Music to Roll Out 'Made With AI' Labels on Tracks
6 hari lalu
Apple Music will require record labels and distributors to tag songs produced with artificial intelligence using a new "Made With AI" label.
Why OpenAI Creates 'ChatGPT for Teens' Feature
8 hari lalu
OpenAI has finally launched a dedicated service for minors, ChatGPT for Teens.
