Deloitte’s 2025 outlook calls this year a ‘pause period’ for generative AI, citing deepfakes, energy demands, and policy gaps.
SYDNEY: Deloitte has labelled 2025 as a “gap year” for the generative AI sector, noting that while interest remains high, the pace of rollout and real-world adoption is showing signs of slowing. The global consultancy released its latest industry outlook, which suggests companies are hitting a period of reflection as they assess technical readiness, ethical standards, and long-term sustainability.
One of the report’s most pressing findings is the potential environmental impact of data centres supporting AI workloads. Deloitte forecasts that AI-related facilities could consume more than 8% of national electricity in developed economies by 2030. This has prompted growing concern about carbon emissions and the urgent need for cleaner energy sources and more efficient computing.
In addition, the report points to a deepening gender divide in AI participation. Women remain underrepresented in both AI education and leadership roles. Without targeted support and inclusion policies, the benefits of AI may remain unevenly distributed.
Deloitte also raises alarms over the rise of deepfakes. The firm notes that synthetic media has advanced rapidly in quality and scale, presenting challenges for online trust, political stability, and content moderation. It calls on technology companies to develop robust detection tools and clear labelling systems to help users distinguish between real and artificial content.
While 2023 and 2024 saw explosive growth in generative AI models and tools, Deloitte suggests that 2025 will act as a stabilisation period. Companies that invest this year in responsible frameworks, transparency, and long-term infrastructure will be best positioned to lead the next wave of AI adoption in 2026 and beyond.
The report concludes by urging industry leaders to treat 2025 not as a slowdown, but as a strategic opportunity to reinforce the foundations of AI innovation.
