FDA accepts AI model to predict lesions in the liver of Absentia Labs – a regulatory framework

Absentia Labs has just achieved an unprecedented achievement: its Digital Liver Model, based on artificial intelligence, was accepted in the FDA's medicine development tools (DDTT) qualification program. It is the first time that an AI tool has received this recognition, paving the way for computer models to complement — and in some cases replace — traditional animal testing and initial clinical trials.

How does the AI model predict liver damage? Operation and data

The system uses neural networks trained in vast liver toxicology databases to simulate human liver reactions to different chemical compounds. Instead of exposing actual guinea pigs or tissues, AI analyzes the molecular structure of the drugs and predicts the likelihood of causing drug-induced liver injury (drug-induced liver injury, DILI). With the acceptance of the FDA's DDT program, drug developers can use the results of this model as evidence in their approval requests, reducing the time and cost of research.

Regulatory framework: why is FDA acceptance historical for AI in health?

The decision signals that the FDA is willing to recognize AI software as reliable tools for security assessment. Until now, the U.S. agency mainly approved AI medical devices, but not predictive models used in the drug discovery phase. The qualification of Absentia Labs can encourage other companies to submit their own solutions, creating a new regulatory standard. In contrast, while Chinese AI advances in the US with low costs that challenge the OpenAI and Anthropic domain, the FDA movement shows that trust in AI for critical areas such as health depends on rigorous validation by regulatory agencies.

Impact on the pharmaceutical industry: how the AI model accelerates testing

In practice, the Digital Liver Model can reduce the need for preclinical tests on animals related to hepatotoxicity by up to 60%, according to company estimates. This not only speeds up the development of new drugs, but also reduces research and development costs, which often exceed $1 billion per approved drug. The approach also promises greater precision, since digital human models can capture genetic and metabolic variations that animal models do not represent.

Future of digital models: AI in health and next regulatory steps

Absentia Labs already works on models of other organs, such as kidney and heart, to expand the platform. FDA acceptance can open doors for other AI tools to be qualified, creating a more agile regulatory validation ecosystem. For the Brazilian market, the news serves as a warning: ANVISA does not yet have an equivalent program, but the American movement can press for local regulatory updates.

Frequently asked questions about the AI model for liver lesions

What does the FDA DDT qualification mean?

It is a program that recognizes tools (such as computational models or biomarkers) as reliable to support drug development, allowing its results to be used in regulatory submissions without further validation by the FDA.

Does the Absentia Labs model replace animal testing?

Not yet completely — the model is used for screening and reducing animal testing, but does not replace all the preclinical steps required by the FDA. The expectation is that, over time, these models become standard and reduce dependence on guinea pigs.

When can the results of this model be used in applications for approval?

From the acceptance of the program, developers can begin to submit data generated by the model as part of their dossiers. However, the FDA may request further validations on a case-by-case basis, until there is more history of use.

> Source: Digital Journal - Absentia Labs Digital Liver Model