ByteDance, the company best known for building the recommendation engine behind TikTok, is now applying artificial intelligence to one of the most difficult challenges in science: drug discovery.
Through its research division Anew Labs, the Chinese technology group has presented early public results involving an AI-designed molecule targeting IL-17, a protein linked to several autoimmune diseases. The development does not mean ByteDance has discovered a ready-to-market treatment, but it does highlight how major technology companies are increasingly entering the pharmaceutical research space, where timelines, costs and scientific complexity remain extremely high.
The significance of the announcement lies in ByteDance’s attempt to use generative AI models to accelerate a process that has traditionally required years of experimentation. In the case presented by Anew Labs, researchers focused on a protein interaction often considered difficult to target with conventional small-molecule drugs. If successful, such an approach could eventually open new possibilities in the treatment of complex autoimmune conditions.
The transition from social media algorithms to biomedical research may appear unexpected, but the underlying logic is similar. Systems such as TikTok’s recommendation engine are designed to process enormous amounts of data, recognize patterns and predict outcomes. Drug discovery involves a completely different scientific context, yet it also depends on navigating vast numbers of molecular combinations to identify those most likely to produce a therapeutic effect.
To support this effort, ByteDance created Anew Labs with operations across China, Singapore and the United States. The company has reportedly developed AnewOmni, a generative AI system trained on millions of biomolecular structures and complexes. The objective is not only to analyze existing biological data, but also to design entirely new candidate molecules, potentially reducing the time required during the earliest stages of pharmaceutical research.
The molecule presented by Anew Labs targets IL-17, a protein associated with diseases including psoriasis, rheumatoid arthritis and ankylosing spondylitis. Current therapies in this area are often injectable biologic drugs that can be expensive and difficult to manage. Developing an effective oral small-molecule alternative would therefore have major clinical and commercial relevance.
Still, caution remains essential. The results released so far are preclinical, meaning they were obtained before human clinical trials. In pharmaceutical development, many promising compounds ultimately fail because of safety, tolerability or efficacy issues that only emerge during advanced testing.
More broadly, ByteDance’s move reflects a larger industry trend. Companies such as DeepMind, Insilico Medicine and Nvidia are increasingly exploring how AI can transform biology and pharmaceutical research. The growing involvement of major technology firms suggests that artificial intelligence is expanding far beyond social media and chatbots into sectors where its long-term impact could be far more significant.
