Overview
Tokenmaxxing — pushing AI usage to its outer limits — was the hottest trend in Silicon Valley earlier this year. CEOs urged employees to “push AI usage as far as it would go,” but cost pressures soon followed.
“Enterprises are still figuring out their AI ROI.” – NEA partner Tiffany Luck, TechCrunch (June 17, 2026)
What Changed
Budget blow‑outs – Uber reportedly exhausted its entire annual AI budget in just a few months.
License reductions – Several firms trimmed Claude licenses for parts of their organization.
Incentive removal – Meta discontinued its internal AI leaderboard, ending a program that gamified model usage.
These moves signal a shift from unchecked experimentation to tighter fiscal oversight.
Analysis
Why It Matters
Focus on ROI – Rapid token consumption is forcing companies to assess AI returns more rigorously, echoing Luck’s observation.
Vendor impact – Cutbacks to Claude licenses suggest that even leading generative‑AI platforms may see fluctuating demand as budgets tighten.
Cultural shift – Dropping leaderboard incentives points to a move away from competition‑driven usage toward measured adoption.
Who Is Affected
Large enterprises – Uber’s budget shortfall illustrates the financial risk of unchecked token spend.
AI service providers – Anthropic’s Claude and similar models could experience variable licensing patterns.
Employees – Teams accustomed to unlimited AI access must adjust to new usage caps and reduced internal recognition.
What to Watch Next
ROI assessments – More enterprises are likely to discuss or publish internal AI ROI calculations.
License restructuring – Companies may renegotiate or consolidate AI tool contracts to align spend with measurable outcomes.
Governance frameworks – Emerging policies around token budgeting and usage tracking are expected to gain traction across the startup ecosystem.
Source
TechCrunch, “NEA’s Tiffany Luck says enterprises are still figuring out their AI ROI,” June 17, 2026, 20:17 UTC.