Earlier this month, Microsoft announced that it had collaborated with Swedish distillery Mackmyra to create the world’s first whisky developed using artificial intelligence (AI). The partnership signals a potentially new wave of innovation whereby craftsmanship becomes displaced by machine learning.

First AI whisky

Whisky making is widely revered as an art, involving a complex understanding of a wide range of ingredients, flavour blends and spirit ageing to create a high-quality final product. However, using its Azure cloud platform, Microsoft used a combination of existing recipes, sales data and customer preferences to generate a dataset of more than 70 million recipes that it predicts will be popular. While the final product, which will be available to purchase later this year, was ultimately tested and approved by a Master Blender (an individual with the specific set of skills required to create new spirit blends and ensure consistencies in existing recipes), the process of creating the recipe was undertaken by a computer.

Automation is already seen as the future of consumer packaged goods across many functions, from e-commerce fulfilment to manufacturing operations to business analytics. However, this AI-created whisky takes automation a step further by taking over a task that is heavily reliant on human sensory systems – particularly the ability to taste and smell. Microsoft insists that the whisky does not replace the expertise and knowledge of a Master Blender, who remains crucial to curating the final product.

Preserving the human element of product creation is particularly vital in a category such as a whisky, where high quality and an interesting brand narrative are fundamental brand components. Furthermore, GlobalData’s 2018 Q3 consumer survey found that one-in-four global consumers consider “high quality” food and drinks to mean those that are “handmade” or “artisan,” underscoring the value of protecting the human connection to the product.

Machines will no doubt open up new avenues for recipe creation across a range of CPG categories, efficiently predicting new ingredient combinations that are poised to succeed based on a wealth of available data. However, even if the final product tastes, smells or looks the same as a “man-made” product, brand perceptions may be irreversibly soured if consumers know that the “artisan” responsible for it is, in fact, a robot.

Latest reports from


Or to search over 50,000 other reports please visit

GlobalData Report Store

GlobalData is this website’s parent business intelligence company.