AI Drives Shifting Costs, Investment and Risk Across Enterprises

Enterprise AI budgets are surging as token costs fall dramatically. The unit cost of inference dropped from about $60 per million tokens in early 2024 to $0.30‑$0.75 by 2026, while average AI spend per Fortune‑500 firm rose from roughly $1.2 M to $7 M annually, with inference now accounting for 80‑90 % of AI compute and budget.

Investment firms report that generative AI tools are expanding the surface area of research, allowing analysts to monitor more companies and generate new insights, though the return on these tools is still being quantified. Business leaders stress that AI adoption must move beyond compliance‑style training to low‑stakes experimentation, with manager visibility boosting perceived AI value by 17‑30 points.

In insurance, experts argue that AI should be applied at the intake stage rather than downstream, where predictive scoring can filter low‑fit risks early, improving underwriting efficiency and reducing wasted effort.

Australian investors face new risks from AI‑driven trading and impulsive app‑based decisions, which may create an “illusion of diversification” and increase portfolio vulnerability.

Overall, the rapid drop in token pricing, rising AI expenditures, and sector‑specific adoption challenges highlight a growing economic and operational pressure point for companies worldwide.