Deloitte has doubled down on its commitment to artificial intelligence with a major new enterprise deal with Anthropic, even as it faces scrutiny over a government report that contained AI-generated inaccuracies. The announcement comes just days after the Australian Department of Employment and Workplace Relations revealed it would require Deloitte to issue a refund for a flawed report produced using AI tools. The report, a A$439,000 “independent assurance review” commissioned by the government, was published earlier this year but later found to include multiple fabricated citations to non-existent academic sources. A corrected version was uploaded to the department’s website, and Deloitte will repay the final installment of the contract, according to the Financial Times. The timing of the announcement is striking—Deloitte revealed on Monday its plan to roll out Anthropic’s AI chatbot, Claude, to nearly 500,000 employees worldwide. The partnership, which began last year, aims to develop compliance-focused AI tools tailored for highly regulated sectors such as financial services, healthcare, and public administration. According to an Anthropic blog post, Deloitte also plans to create AI agent “personas” customized for different roles within the firm, including accountants and software developers. Ranjit Bawa, Deloitte’s global technology and ecosystems alliances leader, emphasized the alignment between the two companies on responsible AI. “Deloitte is making this significant investment in Anthropic’s AI platform because our approach to responsible AI is very aligned, and together we can reshape how enterprises operate over the next decade. Claude continues to be a leading choice for many clients and our own AI transformation,” he wrote. The financial details of the deal, which Anthropic describes as an alliance, were not disclosed. However, it marks the largest enterprise deployment of Anthropic’s AI platform to date, underscoring the growing integration of AI into corporate operations across industries. Deloitte is not alone in grappling with the risks of AI hallucinations. In May, the Chicago Sun-Times retracted an AI-generated summer reading list after readers identified made-up book titles—though the authors listed were real. Amazon’s internal AI productivity tool, Q Business, also faced criticism for inaccuracies during its first year of use. Even Anthropic has faced backlash for relying on AI-generated citations in a legal dispute with music publishers earlier this year, prompting the company’s lawyer to issue an apology. These incidents highlight the ongoing challenges of deploying AI at scale, particularly in high-stakes environments where accuracy and trust are paramount. As companies like Deloitte push forward with AI integration, they must balance innovation with accountability—especially as the technology becomes embedded in everything from government reports to daily business workflows.