{
    "success": true,
    "data": {
        "id": 1729260,
        "msgid": "suspected-indonesian-cracks-grok-using-morse-code-rp-3-4-billion-in-crypto-vanishes-1778390195",
        "date": "2026-05-10 11:15:00",
        "title": "Suspected Indonesian Cracks Grok Using Morse Code, Rp 3.4 Billion in Crypto Vanishes",
        "author": "Soffya Ranti",
        "source": "KOMPAS",
        "tags": "",
        "topic": "Technology",
        "summary": "A suspected Indonesian user exploited xAI's Grok chatbot by embedding hidden Morse code instructions, tricking it into transferring approximately $200,000 (Rp 3.4 billion) worth of DRB cryptocurrency tokens from the Bankrbot system. The attack, which involved sending an NFT to grant additional permissions followed by a deceptive translation request, highlights vulnerabilities in AI agents handling real financial transactions. This incident underscores the risks of prompt injection attacks and has sparked concerns over the security of AI-integrated financial systems, as warned by cybersecurity experts.",
        "content": "<p>A chatbot AI developed by xAI has reportedly been tricked into\ntransferring cryptocurrency assets worth around $200,000 or the\nequivalent of Rp 3.4 billion. This action is suspected to have been\ncarried out by a user from Indonesia, known on the X platform by the\nhandle <span class=\"citation\" data-cites=\"Ilhamrfliansyh\">@Ilhamrfliansyh<\/span>, which is now\ninaccessible. What makes this case intriguing is the method the\nperpetrator used to execute the action, employing hidden messages in\nMorse code to deceive the AI system. The attack was carried out in\nstages. First, the perpetrator sent an NFT titled \u201cBankr Club\nMembership\u201d to Grok\u2019s wallet. This step is believed to have granted the\nAI additional permissions in the Bankrbot system, including the ability\nto conduct cryptocurrency asset transactions. Subsequently, the\nperpetrator asked Grok to translate a seemingly innocuous Morse code\nmessage. However, concealed within it were instructions to transfer\nbillions of DRB tokens to a specific wallet address. Because the system\nregarded the translation result as a legitimate command, Bankrbot\nimmediately executed the transaction, and the assets changed hands. As a\nresult, the perpetrator successfully pocketed around 3 billion DRB\ntokens worth approximately $200,000. Those tokens were immediately sold\non the crypto market, causing a brief fluctuation in the DRB token\nprice. On the social media platform X, several users linked the\nperpetrator\u2019s account to Indonesia based on the language used and\ninteractions in local crypto communities. However, the perpetrator\u2019s\ntrue identity remains unconfirmed to this day. This incident has ignited\nnew concerns regarding the security of AI agents\u2014systems where AI not\nonly answers questions but can also perform real actions such as\nfinancial transactions and digital asset management. A report from the\nEconomic Times states that this incident demonstrates the dangers when\nAI is given direct access to financial systems without adequate\nsafeguards. Grok\u2019s ability to execute commands from translation results\nwithout further verification became the loophole exploited by the\nperpetrator. Cybersecurity experts have long warned of similar threats,\nknown as prompt injection attacks, a technique involving the insertion\nof hidden instructions to manipulate AI behaviour. This case serves as\nreal evidence that such threats are not merely theoretical, especially\nwhen AI is directly connected to financial transactions.<\/p>",
        "url": "https:\/\/jawawa.id\/newsitem\/suspected-indonesian-cracks-grok-using-morse-code-rp-3-4-billion-in-crypto-vanishes-1778390195",
        "image": ""
    },
    "sponsor": "Okusi Associates",
    "sponsor_url": "https:\/\/okusiassociates.com"
}