Global Banks Urged to Be on Full Alert for AI-Related ‘Catastrophe’
Jakarta, CNBC Indonesia - Global regulators are warning of a potential new ‘catastrophe’ stemming from the increasingly massive development of artificial intelligence technology towards full automation. According to a Reuters report on Thursday (11/6/2026), global regulators are calling for the financial system to implement new safeguards amid the accelerating adoption of AI.
The Financial Stability Board (FSB), an international body established by the G20, in its report on Wednesday (10/6) local time, explicitly urged all parties to consider implementing safeguards to mitigate the risks from AI, including from ‘Agentic AI’.
Agentic AI refers to systems capable of planning, reasoning, and executing tasks with limited human oversight. The technology is already being deployed by financial firms to detect fraud, run customer service systems, and manage back-office functions.
As many as 52% of financial sector respondents in a survey conducted by the Cambridge Centre for Alternative Finance reported an active adoption of Agentic AI. A further 23% are transforming and expanding their adoption, while 29% are set to begin using Agentic AI.
Regulators and global standard-setting bodies have been ramping up warnings about the risks posed by AI adoption across the financial sector since Anthropic released Mythos. Experts assess that the advanced AI model presents significant cybersecurity challenges for the banking industry.
The FSB, which acts as the standard-setting body for the global financial system, stated that automated AI triggers major risks, including the possibility of illegal actions, data leaks, and disruptions to interconnected systems.
“AI agents present distinct challenges for human oversight,” the report stated, warning that they could act in ways that diverge from a firm’s intentions without employees’ knowledge or the ability to intervene swiftly.
To address these risks, the standard-setting body has outlined a series of proposed “good practices”. It is urging financial firms to establish clear boundaries on the use of AI and to implement robust safeguards. The non-binding guidelines are open for feedback until 22 July 2026.
The proposals also set limits on what AI agents can do and require human approval for high-risk actions, such as financial transactions above certain thresholds.
The FSB said that financial firms, including banks, could also consider adapting their HR controls and processes towards AI agents by treating them as “synthetic employees”.