Experts have called on organisations to treat artificial intelligence (AI) as critical infrastructure rather than a standalone digital tool, warning that failure to do so could undermine the reliability of key economic and social systems. The call was made in a report published by Punch Newspapers on December 16, 2025, which examined the growing role of AI in modern industry.
According to the experts, AI has evolved beyond experimental deployment and is increasingly embedded in financial markets, national services, and global communication systems. As a result, they argue that AI should be governed and engineered with the same level of rigour applied to power grids, telecommunications networks, and financial clearing systems.
Software engineer Blessing Philips, who featured prominently in the report, noted that public discussions about AI often place excessive emphasis on model development while overlooking the complexity of the systems that support them. She explained that in large-scale environments, the most significant challenges arise not from the algorithms themselves but from the infrastructure required to run them reliably.
News: AI Seen as Key to AfCFTA Success
Philips stressed that effective AI deployment depends on robust, scalable, and resilient infrastructure, including real-time data ingestion pipelines, feature stores that ensure consistency across models, and automated monitoring systems capable of detecting data quality or performance issues before they disrupt operations. She also highlighted the importance of high-performance distributed computing systems that can dynamically scale processing power in response to demand.
In addition, the experts pointed to the need for caching and backup systems to minimise downtime and prevent service outages, particularly in sectors where AI-driven decisions have immediate real-world consequences.
Philips warned that organisations that focus solely on building advanced models while neglecting operational resilience, data integrity, and governance frameworks risk falling behind competitors that prioritise infrastructure reliability. She added that AI failures are increasingly viewed not as technical glitches but as systemic breakdowns with broader economic implications.
The report concludes that the long-term success of AI adoption will depend less on model sophistication and more on the strength, reliability, and scalability of the underlying systems, reinforcing the view that sustainable AI is as much an engineering and governance challenge as it is a technological one.
