Overview

• MIT’s new NANDA study, based on 150 leader interviews, a 350‑employee survey and 300+ projects, finds only 5% lifted sales despite more than $40 billion invested, with budgets overindexed to sales and marketing instead of internal automation.
• AI‑linked tech shares weakened in the second half of August as bubble concerns resurfaced, and OpenAI CEO Sam Altman warned that venture‑backed startups are setting unrealistic expectations that could leave investors with heavy losses.
• Consultancy snapshots show tempered returns: IBM reports only 52% of CEOs see tangible benefits beyond cost cuts and PwC finds 49% expect profitability, while McKinsey highlights scaling hurdles tied to integration, data quality, hallucinations, bias and limited explainability.
• Experts are also questioning sustainability as rising computational demands push energy use and operating costs higher, casting doubt on the economic viability of ever‑larger models.
• In Argentina, many companies keep most data in the cloud yet 44% invest under 1% of revenue in analytics and AI and only 4% treat generative AI as a top priority, even as large turnouts at free training events and school guidance emphasize practical, ethical use.