Pluang Launches Agentic Trading, Allowing Users to Invest via ChatGPT and Gemini
PT Pluang has launched Pluang Agentic Trading, a feature that enables users to directly connect their investment accounts to artificial intelligence assistants such as ChatGPT, Claude, and Gemini for portfolio analysis, investment research, and cross-asset transaction execution within a single conversational session. The feature was introduced on Monday (3/8/2026) and is currently available in a limited beta phase via a whitelist system, with a public launch scheduled for mid-August 2026.
Director of Marketing & Commercial at Pluang, Andreas Agung Hendrawan, stated that the launch represents a step towards providing more equitable access to investment analysis for retail investors across Indonesia. ‘Previously, high-quality investment analysis was a privilege reserved for a select few. Through Pluang Agentic Trading, investors in various regions can now access analytical and execution capabilities on par with those in financial centres,’ Andreas said in an official statement on Monday (3/8/2026).
Andreas explained that agentic trading technology allows AI not only to provide recommendations but also to assist in executing transactions directly on the user’s investment account. However, every transaction must receive manual approval from the user before it is processed. Pluang built this feature on top of Aura AI, its existing AI-based analytical service, with the key difference being that Agentic Trading can execute trade instructions after receiving investor confirmation.
Through this feature, users can conduct real-time portfolio monitoring, buy and sell stocks, gold, and crypto assets, and backtest investment strategies using natural language commands. Pluang claims to be the first regulated retail investment platform in Indonesia to introduce such technology. Globally, similar features were only recently introduced by brokers such as Robinhood, Interactive Brokers, and TradeStation in early 2026.
Pluang Co-Founder Claudia Kolonas highlighted that the main challenge in developing the feature was not enabling AI to place orders, but building a security system to prevent erroneous or unauthorised transactions. The system operates on a propose, confirm, execute mechanism, ensuring the final decision remains with the user. Additional safeguards include daily transaction limits, one-time tokens, and protections against prompt injection attacks. All analysis is provided by the user’s chosen AI model and does not constitute official recommendations from Pluang. The feature operates on Pluang’s infrastructure under the supervision of relevant regulators, with adherence to customer protection standards, Know Your Customer (KYC) processes, and compliance with the Personal Data Protection Law and the Electronic Information and Transactions Law.