Financial Technology Companies Rely on AI to Combat "Deepfake"
For financial technology firms, artificial intelligence can serve to detect deepfake threats and develop virtual assistants. AI has become the heartbeat driving business operations.
Machine learning, or AI, in technology-based financial services is no longer merely supplementary but has become the core driving force of business operations. For financial technology companies, AI can function to detect deepfake threats and develop virtual assistants.
One fintech provider leveraging AI in its operations is PT Akulaku Finance Indonesia (AFI). The company, which focuses on buy now pay later (BNPL) financing services, is also developing AI to detect deepfakes.
Deepfakes are manipulations of images, voice, and video using AI for specific purposes. In the financial services industry, perpetrators often defraud victims by manipulating voice or even images of people close to the victim. Perpetrators typically request victims to send money.
Edy Salim, Director of Information Technology at AFI, stated that deepfakes pose a threat to the digital financial industry, including AFI. “In terms of numbers, I won’t specify, but the possibility (of deepfake incidents) is definitely there,” he said during a media briefing in Jakarta on Thursday (5 March 2026).
For this reason, AFI is developing deepfake AI verification, a tool for detecting manipulation in images, voice, and video. This technology can detect facial image manipulation used by perpetrators. In this way, potential cyber crimes or fraud can be anticipated.
Edy acknowledged that investment in developing deepfake detection tools on AFI’s platform is substantial. “There is no choice (but to invest in this sector). The cost is indeed large, but if not implemented, the risk is even greater,” said Edy, who did not disclose the amount.
According to him, the use of deepfake detection tools is becoming increasingly common among financial technology companies and banks. Especially since, according to VIDA, a provider of electronic certification services offering digital authentication services, deepfake fraud cases continue to surge in the Asia-Pacific region.
According to the VIDA Fraud Intelligence Report 2025, deepfake fraud cases in the region surged by as much as 1,550 per cent. In fact, approximately 97 per cent of businesses in Indonesia have become targets of social engineering or psychological manipulation that exploits human carelessness.
During 2022–2024, according to the report, losses to the banking sector due to digital fraud in Indonesia, including deepfakes, exceeded IDR 2.5 trillion, largely due to the weakness of conventional authentication methods such as SMS OTP and passwords.
For this reason, said Edy, AFI is focused on improving security systems by utilising AI to anticipate deepfakes. “This industry has risks (fraud threats). We, who are in this industry, must mitigate these risks,” he said.
In addition to detecting image, voice, and video manipulation, AFI is also utilising AI to develop credit scoring. This technology, which assesses credit worthiness, is already common in the financial industry. Especially if financial technology companies handle millions of transactions.
“Akulaku Finance was born in the digital era. There is no choice; we must use AI with very large data. In a year, there are over 40 million transactions. If we only use statistical approaches, that is not enough. We must use AI,” explained Edy.
Throughout 2025, AFI recorded new financing worth IDR 7.44 trillion (data prior to audit) with financing disbursement realisation increasing 23 per cent compared to the previous year. Approximately 89 per cent of the entire portfolio comprised BNPL.
President Director of AFI Perry Barman Slangor stated that with credit scoring, the company is able to analyse more deeply regarding user shopping behaviour. In this way, the company can provide credit limits in accordance with the user’s risk profile that has been established.
“Technology is extremely important, indeed the heartbeat for fintech companies such as Akulaku Finance,” said Perry. With a combination of reliable, fast AI technology and a secure ecosystem, he said, AFI can continue to grow healthily and sustainably.
Fundamentally, we must be adaptive in facing these developments, including with technology.
Perry stated that one of the benefits of using AI on credit scoring is the maintenance of non-performing financing (NPF) performance. Last year, for instance, NPF remained at the level of 1.1 per cent. “Year on year, our NPF does not vary much. This demonstrates that our AI is sufficiently reliable and the figure ranges from 1.1–1.2 per cent,” he said.
The company is optimistic that with AI adoption, collaboration with more than 16 banks, and expansion to areas outside Java, AFI’s business will continue to grow. Moreover, this year, the company is targeting new financing reaching IDR 8.2 trillion. “Fundamentally, we must be adaptive in facing these developments, including with technology,” he said.
Another fintech company adopting AI is DANA Indonesia. Since 2024, the digital platform with over 200 million users has collaborated with Microsoft to integrate AI into its business operations and product development.
“Amid the acceleration of digital transformation, AI has become an important catalyst for delivering financial services that are more relevant, adaptive, and secure,” said Ignatia Suwarna, Vice President of Engineering at DANA, in a written statement on Friday (6 March 2026).
One application of AI at DANA is the use of GitHub Copilot as a coding assistant. This solution helps accelerate the application development process whilst improving code quality significantly, with an acceptance rate of 23–30 per cent across various platforms.
AI transformation at DANA also continues with the development of a virtual assistant or chatbot Diana, supported by Azure OpenAI. This service uses an AI agent that can conduct dialogue according to the context of discussion and provide analysis in service development.
The AI agent is claimed to increase productivity.