Deloitte Rolls Out Claude For Global Faces Amidst Scrutiny and Refunds in Australia

Listen to This Article

Deloitte faces backlash after a costly report riddled with AI "hallucinations" forces a refund to the Australian government. Despite it, Deloitte rolls out Anthropic’s AI tech for global operations.

Deloitte has gone all-in on AI, planning a global workforce deployment of Anthropic’s Claude chatbot for nearly half a million employees. At the same time, Deloitte faced a significant setback after errors in an AI-generated report led to a costly refund to the Australian government.

Deloitte’s AI Deployment and Ambitious Plans

Deloitte announced Monday its plans to roll out Anthropic’s Claude to its employees across different countries. The firm is betting big on Claude’s ability to assist with tasks ranging from software development, client engagement and other industry-specific advisory, particularly in regulated sectors like finance and healthcare.

Deloitte’s Ranjit Bawa emphasized how this partnership would reshape operations globally. He said,

"Claude continues to be a leading choice for many clients, and our own AI transformation,"

As part of this AI commitment, Deloitte is investing significant resources into training 15,000 professionals through a formal certification program, ensuring that staff can effectively deploy and maximize the value of AI across their practices.

Deloitte’s Setback in Australia: A Costly Refund

However, the timing of this grand announcement is bittersweet. At almost the same time, Deloitte had to issue a refund to the Australian government after a report, costing A$440,000, was found to be riddled with AI "hallucinations."

These errors, including fabricated citations and inaccurate references, were flagged after being published earlier this year. The report, which was intended to review the Targeted Compliance Framework (TCF), contained flaws such as “system defects” and a lack of traceability in its analysis.

The Australian Financial Review (AFR) and other outlets later reported that the report’s errors were caused by the generative AI tools used by Deloitte, leading to what experts called “hallucinations.” Dr. Christopher Rudge, from the University of Sydney, was one of the first to identify these issues.

Labor Senator Deborah O’Neill criticized Deloitte, calling the situation a “human intelligence problem.” She questioned the over-reliance on AI for critical government reports, suggesting that the refund appeared as “a partial apology for substandard work.”

Deloitte’s Next Steps: Recovery and Innovation

Despite the misstep, Deloitte remains firmly committed to its AI journey. The company plans to roll out Claude’s use across various departments, including accountants and software developers, using specialized AI personas.

A dedicated Claude Center of Excellence will also assist employees in optimizing AI use across teams. This move reflects Deloitte’s ongoing push to modernize its internal operations and support its clients’ AI transformation needs.