Personalized cancer vaccines in 24 hours: China makes a bet that has yet to prove its clinical efficacy

By Momentum IA · June 29, 2026.

Likang Life Sciences has broken ground on what Chinese authorities are presenting as the country's first AI-assisted personalized cancer vaccine manufacturing line. The facility, located in Beijing, has received an investment of around 110 million yuan (16.1 million dollars) and is scheduled to begin operations in October. The stated goal: to cut the design time for an individualized cancer vaccine from weeks to roughly 24 hours.

The process the company describes follows the logic of neoantigen vaccines, one of the most promising lines in modern oncology. A sample of the patient's tumor is extracted, its DNA is sequenced to map the genetic mutations specific to that neoplasm, and with that information a treatment is designed to train the immune system to recognize and destroy precisely those cancer cells. The historic bottleneck has been the analytical phase: processing massive volumes of genomic data required days or weeks of work by specialized researchers. AI algorithms, according to Likang, compress that analysis into a few hours, identifying the most relevant mutations and proposing the vaccine architecture in an automated way.

The figure of 5.15 million new cases of malignant tumors that China recorded in 2024, according to its National Cancer Center, illustrates the scale of the problem this project aims to tackle. It is not just a technological bet: it is also a response to enormous epidemiological pressure. Beijing has strong public-health reasons to accelerate here.

**What the news doesn't say, and it's the most important part**

The announcement is striking, but it's worth reading with surgical precision. That a production line is under construction and that the technology works at clinical scale are two very different claims. The article itself frankly acknowledges that the approach "remains experimental" and that it will have to pass clinical trials and obtain regulatory approval before becoming a widely available treatment. In oncology, that path can take years and has a considerable failure rate, no matter how brilliant the computational design phase may be.

Our reading is that the real value of this announcement is not therapeutic yet, but strategic and infrastructural. China is building the pipeline: the know-how to integrate AI into the personalized vaccine workflow, the manufacturing capacity, the regulatory ecosystem. If clinical trials validate efficacy, that infrastructure will already be ready to scale. If not, it will have learned a great deal with an investment of just 16 million dollars, a modest sum for the pharmaceutical sector.

**A market attracting global capital**

As sector context, the race to integrate AI into drug development is genuinely global and is attracting unprecedented investment. Bank of America estimates that the AI healthcare market could exceed one trillion dollars by 2035. Companies such as BioNTech, Moderna and Gritstone Bio have spent years working on personalized mRNA cancer vaccines, also supported by AI for neoantigen design. The difference with Likang's project lies, for now, in the fact that those companies have published trial results —with promising preliminary results in melanoma and other cancer types— whereas the Chinese proposal has not presented, at least publicly, comparable clinical data.

That does not invalidate the initiative; it simply places it where it belongs: in the phase of demonstrable promise, not consolidated clinical reality.

**The transition no one mentions in the headlines**

There is a dimension that institutional press releases tend to sidestep: the impact on jobs and training that automating genomic analysis entails. The researchers who today spend weeks on that work do not disappear overnight, but their role changes radically. The short term will bring real disruptions in how molecular oncologists and bioinformaticians are trained, which skills become obsolete and which emerge. It is the usual tension of automation: the systemic gain is enormous, but the distribution of that gain is neither automatic nor equitable.

And yet, if the technology works, the long-term horizon is one of the most clearly hopeful that medicine can offer. A vaccine designed in a day and administered within days turns a cancer diagnosis into a fast, personalized intervention, rather than the months-long ordeal it currently represents for millions of people. That is not science fiction: it is the direction the best experimental results point toward, and projects like this one, even if still only promises, build the road that leads there.

The verdict for now is this: a relevant infrastructure milestone, in a field that matters enormously, with a clinical promise that still has to be earned the hard way in trials. It's worth following. With expectations properly calibrated.