👀 Study Reveals How Often AI Gets Financial Advice Wrong 🔹 A study by tech firm Saturn shows that popular chatbots frequently give inaccurate answers to financial questions, reports the FT. Across tests involving AI models ChatGPT, Claude, Copilot, Grok, and Gemini, the models failed 57% of the time on average, with error rates spiking to 88% on complex multi-step calculation prompts. 🔹 The findings highlight the growing risks of relying on generative AI for tax, retirement, and investment decisions, where inaccuracies can cause real monetary losses. 🔹 Researchers evaluated more than 100 money-related questions across free and paid model tiers. In total, 18 models processed over 10,000 queries, with individual prompts repeated up to five times to verify consistency.