Expert: Algorithms Can Be Dangerous, Must Not Be Immune to the Law
The development of digital technology has introduced new challenges to the legal system, particularly regarding the role of algorithms in influencing societal behaviour and information consumption. Conventional legal approaches are deemed insufficient to address this complexity.
Harris Arthur Hedar, Chairman General of the DPN Peradi Professional and Professor at Universitas Negeri Makassar, stated that a shift in perspective is needed when viewing algorithms, which have often been considered neutral and outside the scope of law.
“Technology is never neutral. It carries a mission, design, and consequences,” said Harris on Saturday (18/4/2026).
He explained that the shift in information curation from human roles to algorithmic systems has altered the landscape of information distribution. This condition is seen to imply increased risks of behavioural manipulation and psychological impacts on users.
Harris identified several key challenges in enforcing the law against algorithms, ranging from proving causality, legal status, to cross-border jurisdiction.
“First, legal causality—proving that an algorithm directly causes violence or suicide is difficult. Tech companies will always point to the ‘free will’ of victims or perpetrators as an intervening cause,” said Harris.
He added that from the perspective of legal status, algorithms are not yet recognised as legal subjects, complicating the litigation process. “In the realm of civil lawsuits, class actions require a defendant. Without a juridical construction viewing algorithms as ‘defective products’, victims can only lament their losses without restorative justice,” he explained.
Another challenge arises from jurisdiction, given that many tech companies are based abroad. “Global platforms are often beyond the reach of national laws in developing countries, so even if there is intent to sue, executing the ruling becomes utopian,” he stated.
According to Harris, current legal approaches still rely on the concept of clear legal subjects, such as companies in conventional products. Meanwhile, algorithms are considered more complex because they are dynamic and intangible.
He exemplified the difference with physical products. “If cigarettes cause cancer, we sue the cigarette company. If cosmetics cause skin damage, we sue the cosmetics producer. However, algorithms are different. They are not goods in the classical sense, but systems that continuously evolve,” he said.
For this reason, he proposed several approaches, including expanding the concept of gross negligence in civil law and reconstructing algorithms as products within the framework of product liability.
“If a platform knows that its algorithmic design has the potential to trigger polarisation or violence but still prioritises engagement for profit, that can be categorised as negligence,” Harris clarified.
He emphasised that efforts to sue algorithms are not intended to hinder innovation, but to ensure legal protection continues in the digital era.
“It is time for the law to act as the commander in cyberspace, ensuring that technological innovation aligns with human dignity and values of justice,” he said.