According to a Morgan Stanley report, AI data centers are set to dramatically increase global water consumption over the next few years. By 2028, these facilities may utilize approximately 1,068 billion liters of water annually, an 11-fold jump from 2024 levels.
That estimate draws from three components:
- On-site cooling of data centers (Scope 1)
- Electricity generation required to power centers (Scope 2)
- Semiconductor manufacturing (Scope 3)
Depending on efficiency improvements, technologies adopted, and the regional energy mix, the report estimates a possible range from 637 billion to 1,485 billion liters annually by 2028.
Why This Matters
- Environmental Risks: Over half of existing data center hubs are located in regions already facing drought, flooding, or declining water quality—raising the alarm on resource stress.
- Water Scarcity Pressure: AI’s projected water demand—comparable to 6.6 billion cubic meters by 2027—mirrors the annual water consumption of multiple countries such as Denmark or half of the UK.
- Investor & Regulatory Scrutiny: Growing concern is prompting investors and global regulators to begin monitoring water use alongside energy and carbon metrics in tech infrastructure. Investopedia
- Sustainability in Spotlight: In response, innovations like water-efficient cooling systems, water recycling, and regulation of data-center placement are gaining traction.
Summary at a Glance
| Metric or Insight | Figure or Context |
|---|---|
| Projected Annual Water Use (2028) | ~1,068 billion liters (base case) |
| Possible Range by 2028 | Between 637 – 1,485 billion liters |
| Key Drivers | Cooling, electricity generation, chip production |
| Regional Risks | Many hubs in water-stressed areas |
| Broader Impact | Comparable to annual withdrawals of small countries |
| Sustainability Trends | Water reuse, efficient cooling, regulatory action |
Concluding Thought
The meteoric rise in AI is powering a parallel surge in water demand—with data suggesting global consumption could exceed 1 trillion liters annually by 2028. As AI scales globally, sustainable water management will be a critical factor in balancing technological advancement with environmental stewardship.
