Generate Biomedicines announced on October 13, 2026 that it has signed a multi-target collaboration and license agreement with Danish drugmaker Zealand Pharma to design multispecific protein therapeutics for obesity and other metabolic diseases, using Generate’s generative AI platform to invent entirely new protein structures rather than screen existing compound libraries.
The Numbers Behind the Deal
Generate will receive $30 million upfront and up to $35 million in near-term preclinical milestones for the two initial programs, with total potential milestone payments across the partnership reaching nearly $1.2 billion plus tiered royalties on worldwide sales. The deal also includes an option for Zealand to add a third program later, meaning the final value could climb further if early results hold up.
A New Modality for Zealand
Zealand Pharma has built its pipeline almost entirely on peptide-based medicines, the same broad drug class behind blockbuster GLP-1 obesity treatments. Chief scientific officer Uptal Singh framed the Generate tie-up as an expansion of that toolbox to include proteins, specifically multispecific proteins engineered to hit several biological targets at once rather than the single-target approach typical of most peptide drugs. That could let Zealand pursue metabolic conditions where hitting one pathway alone has produced diminishing returns.
How Generate’s AI Platform Works
Generate Biomedicines, a Flagship Pioneering-backed company that went public earlier this year, uses generative AI models to design novel protein sequences and structures computationally before any wet-lab synthesis begins, compressing a process that traditionally relies on screening naturally occurring or previously catalogued proteins. Under the new agreement, Generate will lead research through lead-candidate selection, after which Zealand takes over clinical development, regulatory submissions, and global commercialization — a division of labor that lets the AI-native company focus purely on design while leaning on Zealand’s existing regulatory and commercial infrastructure.
Part of a Broader Pattern of Billion-Dollar AI Bets
The Zealand deal adds to a string of large partnerships Generate has struck with major pharmaceutical players, including prior billion-dollar-plus alliances with Novartis and Amgen. It also lands amid a broader surge in AI-pharma licensing: DealForma data cited by industry trackers put AI-based drug discovery deal volume at $47.1 billion in the first half of 2026 alone, already surpassing the $44.6 billion recorded for all of 2025. Obesity and metabolic disease have become one of the most contested fronts in that spending, given the commercial success of existing GLP-1 drugs and pressure on manufacturers to find next-generation alternatives.
Reasons for Caution
Skeptics of the AI-protein-design boom note that no multispecific protein designed primarily by a generative AI platform has yet reached FDA approval, and that headline milestone figures like Zealand’s nearly $1.2 billion are back-loaded, contingent on years of preclinical and clinical success that most drug candidates never achieve. The companies have also not disclosed the specific biological targets involved, making it difficult for outside observers to assess how differentiated the resulting molecules might be from existing obesity treatments already on the market or in late-stage trials.
The Optimistic Case and What Comes Next
Supporters argue that generative protein design is maturing faster than generative molecule design for small-molecule drugs, since proteins are themselves the output format AI models like Generate’s were built to predict, potentially shortening the path from computational design to a dosable candidate. The next milestones to watch will be whether Generate nominates a lead candidate within the deal’s early preclinical window, and whether Zealand exercises its option to add a third program — a signal of how much confidence the company places in the AI-designed proteins before any have been tested in humans.
Photo: Pavel Danilyuk / PEXELS via Pexels
