As artificial intelligence (AI) continues to gain traction, brands increasingly face the question of its environmental impact. AI technologies, particularly large models like GPT-3, require immense computational power, consuming significant energy in data centers. Research by Alex de Vries indicates that AI servers could use as much electricity as a small country annually by 2027. This energy demand understandably raises concerns among sustainability-minded retailers and consumers alike.
However, AI’s energy consumption must be viewed in context. Essentially, every business function exerts some environmental impact, often exceeding that of AI alone. For instance, PwC’s internal studies show that emissions from embracing generative AI amounted to only a fraction of what business travel typically produces. Therefore, while energy use is critical, brands should weigh AI’s advantages—like improved demand forecasting and reduced overproduction—against its energy costs.
Additionally, comparisons of AI-generated tasks to human labor reveal that AI can often yield lower emissions due to its efficiency. Research from the University of California suggests that using AI for writing and imaging tasks can produce fewer emissions than human efforts when accounting for task duration and the corresponding energy consumption.
As brands assess the environmental implications of AI, they must also consider their overall operational impacts. While AI will contribute to ongoing electricity demands across sectors, effective strategies to mitigate emissions should take priority. Ultimately, contemplating AI’s role in their sustainability efforts will determine whether brands benefit or bear costs associated with its environmental footprint.