Heaptalk, Jakarta — The rapid development of AI has brought several consequences, one of which is ethical concerns. At the Forum Digital BUMN (Fordigi) Summit 2024, practitioners and academics discussed ethical AI practices applied by multiple companies in Indonesia, both state-owned enterprises and private companies.
According to Prof. Eko Kuswardono Budiardjo, a professor at the Faculty of Computer Science, University of Indonesia, the country already has a national strategy for artificial intelligence from 2020 to 2045 as a framework for national policy. Humans play a central role in utilizing AI technology, particularly regarding human agency and oversight. This role can be carried out through human-in-the-loop, human-on-the-loop, and human-in-command.
“We shouldn’t be afraid of AI since AI also adheres to what is known as ethical AI, which includes governance, bias, and transparency. Therefore, according to the challenges, we can use AI technology appropriately,” Prof. Eko concluded.
Ensuring data safety and security
Kurnia Sofia Rosyada, Group Head/SVP of Enterprise Data Analytics at Bank Mandiri, highlighted that the evolution of generative AI is evidence of the rapid advancement of technology. The use of large language model (LLM) applications and the trend of GPU-based computing is expected to rise along with the increasing adoption of AI.
Investment in computing power and sufficient electrical energy will become essential to support AI development. “AI is here to stay. It depends on how we utilize AI—whether in a way that benefits us or creates challenges or difficulties for us,” Kurnia said.
In addition, Kurnia emphasized the importance of ethical AI, particularly in ensuring data safety and security. She added, “How do we ensure that with the implementation of the Personal Data Protection Law, which will be effective in October 2024, the privacy rights of our customers are maintained? Then, there is transparency and explainability in the AI models themselves. Equally important are awareness and literacy.”
On this occasion, Herry Prajitno Harjono, Head of Fraud and Financial Crime at SAS ASEAN & Korea, explained the application of trustworthy AI to balance responsibility and profitability. Implementing trustworthy AI has four pillars: oversight, operations, compliance, and culture. “The use of AI is not solely about benefiting certain interests, but also about balancing the interests of other stakeholders, particularly customers,” Herry concluded.
