AI Accountability Crisis: Australian Firms Can't Explain AI

Many Australian organizations using AI in financial processes, including insurance, can't properly explain their AI agents' actions to regulators. What's more, nearly one in five can't even identify who would be held accountable if something goes wrong. This comes as regulators and industry bodies note that governance isn't keeping up with AI deployment. A report found that 18% of respondents said accountability for a significant AI error would be unclear or non-existent. Another 18% believe the executive who approved the AI investment would be personally responsible, showing a lack of formal structures. A skills gap also exists, with 75% of respondents lacking in-house expertise to understand their AI agents. Only 59% are confident they can explain an AI agent's actions to an auditor. This falls short of current regulatory requirements. The pressure to deploy AI agents quickly is high, with 59% prioritizing speed over governance. This situation highlights a growing risk in the rapid adoption of AI without clear accountability.