2 weeks ago
Goldman Sachs: US firms ramp AI spending, earnings impact limited
A big bank called Goldman Sachs studied how American companies are using artificial intelligence, or AI.
They looked at how much money companies earned in the spring of 2026.
The companies that build AI tools made a lot more money than before.
But most other companies have not seen their profits grow much from using AI yet.
Only a few companies could say exactly how much AI helped them.
Many companies are spending more money on AI every month.
In July, the average company spent $12 per worker on AI, up from earlier in the year.
This spending is still tiny compared to what companies earn overall.
Most companies are paying for AI by moving money from other budgets.
Experts think the benefits might become clearer as companies use AI more widely.
S&P 500 Q2 2026 earnings growth is tracking at 31% year-on-year, excluding 'other income' linked to private investment stakes.
Earnings of hyperscalers and AI infrastructure companies rose 54% year-on-year, accounting for about half of S&P 500 earnings growth.
Just 11% of S&P 500 companies quantified AI productivity gains for a specific use case, and only 2% quantified AI's impact on earnings.
Median monthly AI spending per employee rose to $12 in July from $5 at the start of the year; top 10% of spenders reached $650 per employee.
AI inference expenses represent less than 0.5% of S&P 500 revenues, with about two-thirds of companies funding AI by reallocating existing budgets.
- Who
- Goldman Sachs analysts and S&P 500 companies
- What
- AI spending is rising sharply across US firms, but the direct earnings impact of AI adoption remains limited
- Where
- United States
- When
- Analysis of Q2 2026 earnings, reported on August 16, 2026
- Why
- Companies are moving AI from experimentation to wider deployment, while measurable productivity gains are still at an early stage
Key facts
- Publisher
- Goldman Sachs
- S&P 500 earnings growth (Q2 2026)
- 31% year-on-year (excluding 'other income')
- Hyperscaler/AI infrastructure earnings growth
- 54% year-on-year
- Companies quantifying AI productivity gains
- 11% of S&P 500
- Companies quantifying AI earnings impact
- 2% of S&P 500
- Median monthly AI spend per employee (July)
- $12 (up from $5 at start of year)
- Top 10% AI spenders per employee
- $650 (up from $240)
- AI inference share of S&P 500 revenues
- Less than 0.5%









