In fashion, AI has first been embraced by big brands such as Amazon, Adidas, and Zara. Adidas, for instance, has voice-assisted in-store robots, while competitors are now following Zara after it partnered with Jetlore for a consumer behavior prediction platform. Zara’s aim to create “full integration between store and online stock rooms” is becoming an industry trend. Across the fashion world, consumer-facing AI solutions have been focused on improving personalisation and customer experience.

But AI has also crept into supply chain optimisation, inventory management, and other aspects of production. H&M and Adidas have begun integrating ‘smart’, interconnected robots into production to gain leverage. On the other side of the room are efforts to predict fashion trends and create products that fit these trends. For instance, Indonesian entrepreneur Lingga Madu has gained global attention after using AI to predict what styles will blow up and which ones will flop.

Similar efforts can be seen in the cosmetics industry, as the sector is now tapping into AI to improve customer engagement and relevancy. Mobile apps such as Modiface and HiMirror claim to be capable of assessing your face’s quality and even give you skincare advice. Estée Lauder, Sephora, Benefit, and others have developed their own apps with basically the same features, all powered by AI. Olay’s wildly popular app even lets you identify which parts of your face are most prone to ageing while also addressing other problems.

Taking this concept further, Swedish beauty tech brand FOREO’s has developed a cleansing device called LUNA fofo which can read hydration levels and suggest a personalised skincare routine. FOREO founder and CEO Filip Sedic even said that they plan to “detect air quality and [the] user’s skin conditions in real time,” making the device “the world’s smallest beauty coach.”

Nonetheless, despite these lofty claims, these fashion and beauty innovations are not without some blind spots. While AI has the capacity to process huge amounts of skin data based on user photos and facial recognition, factors such as lighting and picture quality play a huge part in the overall assessment. Indeed, there is a lot of pseudo-science in the world of skincare. For instance, Olay’s vague concept of “skin age”, while its very convenient idea, it is not backed by hard science.

While we can’t doubt that AI and machine learning can transform these industries, it is difficult to make out the exact shape of the things to come. When asked about using AI to address negative fast fashion practices, Elle writer Jo Fuertes stated that a sustainable and ethical future for fashion “requires a seismic cultural shift in how humans work together, not a technological one.” AI is driving more personalised consumer experiences, but it is still underpinned by traditional market models. Unless we see AI radically transforming these models from the ground up, what we can only expect from it today are on-point recommendations and fancy camera tricks for your social media feed.