3 Minutes
Businesses Betting Big on Generative AI
Global organizations are accelerating investment in generative AI (GenAI), often prioritizing innovation and competitive edge over environmental and operational consequences. A recent industry study finds 61% of firms plan to boost GenAI budgets in the next 12 months, following a year when 88% increased spending on AI technologies. That rapid expansion highlights growing enterprise AI adoption — but it also raises urgent questions about sustainability, cost control and long-term ROI.
Energy Use and Sustainability Concerns
While attention has focused on AI platform providers, end-user companies are driving significant demand that translates into higher data center energy consumption. Major cloud operators have reported steep increases in energy use: one provider recorded a near 27% uptick in data center power during parts of 2024–25, while another saw overall energy consumption grow dramatically over recent years. Despite these indicators, more than half of surveyed companies believe GenAI’s benefits outweigh its environmental impact — a view especially common in some regions.
AI carbon footprint and hidden costs
Beyond sustainability, firms face unexpected expenses as AI projects scale. Around one in five organizations report dissatisfaction with early outcomes and frequent “bill shock” when resource usage outpaces forecasts. That underscores the need for cost modeling, observability and governance before broad deployment.
Product Features and Technical Choices
Enterprises are exploring a variety of AI models and deployment patterns to balance performance, cost and privacy:
- Large foundation models for advanced capabilities and high-quality generative outputs.
- Small language models (SLMs) for edge deployments, lower inference costs and faster response times.
- On-premises or private cloud hosting to control data flows and meet compliance requirements.
- Hybrid inference architectures that route sensitive queries to secure environments while using public clouds for less critical workloads.
Comparisons: Large Models vs Small Language Models
Large models typically deliver superior contextual understanding and creative generation but come with higher compute, energy and operational costs. Small language models trade some capability for dramatic savings in latency, inference costs and carbon emissions — making them attractive for customer-facing applications and regulated industries.
Advantages, Use Cases and Market Relevance
Strategic GenAI adoption delivers clear advantages: improved customer operations, automated marketing personalization, faster risk modeling and enhanced decision support. Use cases include chatbots and virtual assistants, intelligent document processing, personalized content generation, and fraud detection. The market relevance is evident as organizations allocate an average of about 12% of IT budgets specifically to generative AI initiatives.
How to Maximize ROI and Minimize Impact
Spending alone isn't enough. To get measurable returns, enterprises should build a strong data foundation, enforce AI governance, prioritize secure and compliant environments, and focus pilots on high-impact domains like customer operations, marketing and risk management. Yet only a minority have formal governance in place today, leaving many businesses exposed to cost overruns and regulatory risk.
In short, the next phase of enterprise AI should pair ambitious investment with disciplined engineering, sustainability planning and governance to ensure generative AI scales responsibly and delivers durable business value.
Comments