We all know how AI is revolutionizing our lives, our work, and the emergence of new startups looking to ride the wave of this trend. A sector that might seem unrelated to this world at first glance is biotech. However, the reality is quite different...
In recent years, artificial intelligence (AI) has taken center stage in the biotech sector, attracting massive investments and raising high expectations about its potential to transform drug discovery and development. One of the latest examples confirming this trend comes from Retro Biosciences, a startup backed by Sam Altman, CEO of OpenAI, who, after providing an initial $180 million investment, is now further contributing to a Series A round of approximately $1 billion.
In partnership with OpenAI, Retro Biosciences has developed a custom AI model that designs proteins capable of temporarily converting normal cells into stem cells, with the goal of reversing the aging process.
This example highlights the growing visibility of the sector and the strong interest of investors in AI-driven biotech. Companies like Insilico Medicine, Recursion, BenevolentAI, and Exscientia are advancing predictive AI-based platforms to discover new drugs and therapeutic targets. However, the industry is still in a phase of adjustment. The recent clinical failures of companies such as Calico, Neumora, and BioAge Labs underscore the challenges of translating AI-driven promises into real-world successes.
Moreover, the primary focus today is no longer just on algorithms but on the quality and quantity of data. The real competitive advantage now lies in access to proprietary, well-structured datasets, leading the industry into a "data race" rather than an "AI race." Leading companies are building infrastructures to acquire and integrate vast amounts of biological data, while major pharmaceutical firms are forming strategic partnerships to maximize the potential of these technologies. The future will be shaped by the ability of these companies to integrate AI into R&D processes with a pragmatic approach, balancing innovation with experimental validation.
Challenges and How startups are evolving
Earlier, we provided a small piece of the puzzle that, in our view, clearly demonstrates the traction this sector is gaining and the growing attention it’s receiving from key stakeholders in the Life Sciences industry. The potential revolutions in R&D mainly revolve around two aspects: reducing the time and capital required to bring a drug to market. While the benefits of AI technology in drug discovery are relatively clear, for VC investors, the real challenge lies in identifying which startups have the "secret recipe" for success. Traditionally, biotech startups focused on developing cutting-edge treatments with one or two products in their pipeline, with the primary goal of selling their assets upon reaching clinical proof of concept (typically Phase 2 trials).
In recent years, however, the landscape of biotech startups integrating AI into drug discovery has been changing dramatically. Whereas these companies were once structured around a small core of scientists and managers from the pharmaceutical industry, today we see the emergence of much more diverse teams. It is no longer just biologists and chemists driving innovation, but also machine learning specialists, entrepreneurs with tech backgrounds, and even data management experts. This mix of expertise has become essential to tackle the increasing complexity of AI-driven drug discovery.
The growing financial needs
This evolution in the business model has a direct impact on the financial needs of startups. Whereas in the past, the capital raised was primarily allocated to R&D and clinical trials, today the funding requirements have shifted significantly. Companies must invest heavily in building advanced technological platforms, acquiring and generating high-quality data, and assembling teams dedicated to managing commercial partnerships. This explains why funding rounds are becoming increasingly larger and why some industry players are receiving extremely high valuations, aiming to establish themselves as true leaders at the intersection of AI and biotechnology.
The ritical Challenge: Data access
The key factor determining the success of AI-driven biotech startups will be their ability to acquire and leverage high-quality data. AI in drug discovery has already demonstrated its potential, but without robust, specific, and well-structured datasets, even the most advanced platforms risk generating unreliable results. Today, the real race is not so much about developing the most sophisticated algorithm but about building the best data infrastructure. One of the biggest challenges in this sector, compared to other industries where AI is applied, is the limited amount of training data available for models. While other industries can rely on datasets containing millions or even billions of samples, clinical studies, for example, often have only a few dozen patients. This presents a significant challenge for developing reliable predictive models.
Startups that can successfully collect, generate, and safeguard proprietary data will not only gain a decisive competitive advantage but will also become the most sought-after partners for investors and big pharma companies.
A New Business Model: No longer just proof of concept
AI-driven biotech startups are also evolving in their business models. In the past, the primary goal was to achieve a clinical proof of concept (POC) in Phase I-II to demonstrate the validity of a drug candidate and attract the interest of big pharma. Today, however, companies are focusing from the outset on building broader product pipelines and establishing strategic co-development agreements with pharmaceutical companies, with two main objectives: gaining commercial traction and diversifying risk across multiple assets.
The pharmaceutical industry is now a well-established market for AI partnerships: it has the necessary capital and knows exactly what it wants. Moreover, it has already extracted value from these collaborations. In this sense, pharma represents the ideal partner. The creation of proprietary AI platforms is therefore seen as a long-term asset, capable of generating value not only through the development of in-house drugs but also through collaborations and licensing agreements.
Exit Scenarios
Finally, exit scenarios for AI-driven biotech startups are evolving, with two main emerging pathways: IPO and Buy and Build. The IPO represents the most ambitious option, as seen with Exscientia, which went public on Nasdaq, raising over $500 million to develop its drug discovery platform.
In this new competitive landscape, more Life Science startups will likely find their ideal exit through buy-and-build strategies, where larger companies acquire them to strengthen their capabilities and integrate data/algorithms. A notable example is Recursion Pharmaceuticals, which acquired Cyclica and Valence Discovery to expand its computational pipeline, or Valo Health, which integrated TARA Biosystems to enhance its cardiovascular research.
This approach allows companies to consolidate expertise and data, accelerating drug development and creating more competitive players in the industry.
Conclusions
As IAG, we are witnessing the growing development and consolidation of these businesses, both at the European and international levels. One of our portfolio companies, Rubedo Life Science, is a prime example of this trend.
Thanks to its proprietary AI-driven drug discovery platform, ALEMBIC™, which combines sophisticated computational algorithms with chemistry, the startup can generate numerous drug candidates. These candidates will be studied in preclinical stages to strengthen their pipeline and establish new partnerships. Confirming this strategy, the Italy-California-based startup announced a partnership with Beiersdorf in 2024 for the identification and development of molecules in the dermatology field. As we previously analyzed, the team is internally conducting Phase I clinical trials to assess the efficacy and safety of treatments for dermatological conditions.
At the same time, potential drugs for respiratory, metabolic, and neurological indications are being developed to diversify and attract new co-development agreements. Rubedo, like many of the examples we have discussed, is just one of many companies shaping this new landscape—it won’t be the first, and it certainly won’t be the last.
We are entering what we like to call a new era for biotech startups. Stay tuned for what’s coming next.