Eskom to Rationalise AI Pilots as Costs Rise

Why It Matters

Eskom’s AI cutback highlights the tension between innovation and fiscal discipline in utilities, signalling that large‑scale AI rollouts must deliver measurable value to survive. The decision could set a benchmark for other emerging‑market power providers facing similar cost pressures.

Key Takeaways

• Eskom runs 220 AI pilots, seeking cuts
• Business case required for each AI project
• Focus on predictive maintenance, grid intelligence, digital twins
• Liberalisation drives need for dynamic pricing AI
• Mirrors J&J pilot reduction, avoiding waste

Pulse Analysis

Eskom’s decision to prune its AI pilot programme reflects a broader shift in the utility sector toward disciplined digital transformation. While AI promises to unlock efficiencies in asset management and grid operations, the capital‑intensive nature of pilot projects forces companies to scrutinise spend against tangible outcomes. By demanding rigorous business cases for each initiative, Eskom is attempting to safeguard its balance sheet amid rising operational costs and the fiscal scrutiny that accompanies its status as a state‑owned enterprise.

The utility’s AI focus areas—predictive maintenance, intelligent grid control, and digital twins—are directly tied to South Africa’s ongoing electricity‑sector liberalisation. As independent power producers enter the market, Eskom must manage a more fragmented supply landscape, dynamic pricing structures, and complex wheeling arrangements. AI can process vast streams of sensor data, forecast equipment failures, and optimise power flows, thereby enhancing reliability while curbing fuel and maintenance expenses. These capabilities are critical for meeting regulatory targets and maintaining competitive tariffs in a deregulated environment.

Eskom’s approach mirrors the experience of multinational firms like Johnson & Johnson, which slashed 700 of its 900 AI pilots to eliminate waste. The lesson is clear: without clear ROI metrics, AI initiatives can become costly experiments rather than strategic assets. By narrowing its focus to high‑impact projects, Eskom not only reduces immediate spend but also builds a governance framework that other utilities in emerging markets may emulate. Successful execution could position Eskom as a model for cost‑effective AI integration, driving sector‑wide improvements in efficiency and service reliability.