MIT survey exposes AI's sustainability measurement gap
Efficiency gains coexist with rising resource use in automated warehouses.
AI's reported sustainability benefits are not consistently reflected in total emissions. MIT News says 87% of adopters reported positive effects on sustainability goals, but 29% of mature agentic-AI users reported higher total greenhouse-gas emissions.
Companies with cross-functional sustainability teams embedded sustainability in decisions more often: 57%, compared with 13% under single-function responsibility. Sustainability was important to long-term success for 83%; only 33% of companies with public goals had board-level oversight. Only 51% had a formal cross-functional team.
What the survey can establish
The underlying report surveyed 1,810 professionals in 91 countries between February and April 2026. The voluntary survey used purposive and snowball sampling, with different role-based question pathways. These are participating professionals' accounts of their organisations, rather than a census of global businesses. The questionnaire offered five languages and averaged forty-five questions.
Among advanced or fully automated warehouses, 78% reported higher energy use after AI implementation and 76% higher waste. The report's 24% real-time AI/sensor monitoring figure concerns warehouses overall, rather than that automation subgroup. The survey cannot separate AI's effect from higher throughput at rapidly growing warehouses.