Ahmad Ramli, a professor of cyber law at the Faculty of Law at Padjadjaran University, is pushing for transparency in digital platform algorithms to ensure fair content distribution and prevent allegations of reach restrictions or shadowbans that are difficult for users to verify.
Hot Topics
This reflects the accusations against the American multinational technology company, Meta, in the context of the recent presidential election in Brazil.
"This is certainly urgent, especially during political moments like the approaching and ongoing elections. But it's important to understand that the term 'opening up' doesn't necessarily mean giving all of Meta's secrets to the public," Ramli said.
He explained that transparency does not have to be achieved by revealing all of a company's source code or trade secrets.
Platforms may provide limited access to independent auditors or regulators, along with public reports containing key indicators regarding content distribution and restrictions.
According to Ramli, one of the main problems is the difficulty of determining whether or not algorithmic bias exists when platforms do not provide adequate data.
This situation can give rise to suspicions that certain content or accounts receive lower reach than others without a testable explanation.
"The problem is that it's difficult to verify whether bias exists. This is where transparency is important," he said.
To address these issues, Ramli is pushing for regular algorithm evaluations, including examining the error rates of account blocking across various groups and providing open explanations of the factors that influence content recommendations and reach.
He also believes that automated moderation decisions should be accompanied by a human review mechanism, whenever technically feasible. Platforms should provide a swift redressal pathway for users who feel aggrieved by erroneous restrictions or blocks.
Tags:
asiaeast


