Ten years ago, a ban on animal testing for cosmetic products came into force in Europe. Ethics and regulatory requirements inspired cosmetic industry to develop alternative models. In vitro approaches are widely used in cosmetic safety tests. Skin/eye irritation, sensitization, phototoxicity, genotoxicity tests have been validated by the European Centre for the Validation of Alternative Methods (ECVAM) and approved by the Organization for Economic Co-operation and Development (OECD). It seems like major obstacles have been successfully overtaken. Nevertheless, questions about systemic toxicity, developmental and reproductive toxicity are still open. Two-dimensional cell culture models are not able to precisely describe the whole complexity of in vivo skin health responses on a toxic ingredient. Single cell type cannot represent a multicellular organ, let alone an organism. That is why an integrated testing strategy is a popular choice to cover the variety of possible negative effects of test products. Development of three-dimensional models of human skin, like Reconstructed human epidermis, was an important breakthrough. They have a close structural and physiological resemblance with human epidermis, proposing also an open-air surface for topical applications. Still, absence of vascularization (epithelial cells), innervation (neurons), immune response (Langerhans cells), pigmentation (melanocytes), skin appendages (hairs and glands) distances this model from reality. Despite the mentioned limits, the currently available alternative models and validated tests have enough predictive power to reliably describe toxic effects of cosmetic products and ingredients.

New generation of 3D-skin models and “organ-on-chip” approach look promising. They may guide us to a new level of safety tests or, at least, expand our investigative capabilities. Advanced tissue engineering allows to produce full-skin models incorporating different cell types. These models are close the real human skin, both structurally and physiologically. Microfluidic-based organ-on-chip device is a promising choice when a simultaneous reply to a treatment from various cell types is needed. A compact chip allows to cultivate various cells in different compartments, mimicking function of an organ, while microfluidics provides precise control of the environment. However, these approaches are not accepted yet by the official authorities.

Ban on animal tests impacted also efficacy tests. In vitro cell-based approaches are not adapted for final cosmetic formulations, while popularity of 3D-models in evaluation of comprehensive biological effects of final cosmetic products or even active ingredients is low. Efficacy tests are mainly performed on living human skin explants or on volunteers in clinical studies.

In silico approaches became popular for safety prediction in cosmetic industry. Large amount of data about toxicity of chemicals is available now. A recent publication identified over 900 databases, covering biology, drug discovery, chemistry, omics, clinical trials, toxicology etc (1). Only in silico tools are able to compile this data and use it for predictive toxicology of cosmetics.

Prediction is based on the structure of target molecule. Logically, similar chemicals are expected to cause similar effects. However, reality is more sophisticated and tiny changes in chemical structure may change its properties.

Discussion about perspectives of AI in cosmetic testing would not be complete without participation of AI itself. Here is the answer of the most popular AI nowadays to the same questions:

The cosmetics industry faces several obstacles in testing, particularly with the growing bans on animal testing. However, companies are actively exploring alternative methods to address these challenges. Here are some key obstacles and the corresponding alternatives being explored: