AI drug discovery is moving beyond laboratory pilots into real clinical programs, where AI agents trained on actual regulatory constraints and patient data are now helping researchers navigate the full pathway from target identification to patient treatment. Lantern Pharma's Open-Medicine AI (OMAI), a wholly owned subsidiary, is preparing to unveil how this approach works at scale, with plans to present its multi-agentic platform to investors and industry partners on September 23, 2026.

What Makes AI Built Inside Clinical Programs Different?

Most AI tools for drug discovery are developed in isolation, trained on historical data or theoretical models without exposure to the messy realities of actual drug development. Open-Medicine AI takes a fundamentally different approach. The platform's AI co-scientists were developed within Lantern's own clinical-stage oncology programs, meaning they learned from the same data constraints, regulatory requirements, and decision points that determine whether a molecule actually reaches patients.

This distinction matters because drug development involves far more than identifying promising compounds. The process spans target and biomarker selection, compound and modality design, translational analysis, trial design, and regulatory strategy. Open-Medicine AI's agents are designed to reason across this entire pathway, carrying the operating knowledge of the scientists and clinicians who built them.

The platform was first commercialized as withZeta.ai, launched in April 2026 and focused on rare and aggressive cancers. In July 2026, the company expanded it with ZetaOmics, a computational biology module released to early-access users. Now, Open-Medicine AI is extending this architecture beyond oncology into additional disease categories, new therapeutic modalities, and the downstream challenges of drug development that consume most of the time and cost between a validated target and a treated patient.

Why Is the Market Moving Toward Multi-Agent AI Systems?

The shift toward AI platforms like Open-Medicine AI reflects broader industry trends. Independent analysts estimate the AI in drug discovery market at approximately 2.9 billion dollars in 2026, growing to 13.8 billion dollars by 2033. Additionally, analysts have noted that AI deployments in pharma are moving from pilots to enterprise-scale, agentic systems.

This growth is being driven by real clinical success. The 2026 Lasker Awards, announced in September 2026, honored three major advances in drug discovery and treatment that demonstrate how foundational research translates into patient benefit. These awards underscore the value of combining rigorous science with practical development approaches.

How to Evaluate AI Drug Discovery Platforms for Your Organization

• Training Data Source: Determine whether the AI agents were trained on historical data alone or within active clinical programs with real regulatory and patient constraints, as this affects their ability to navigate real-world development challenges.
• Scope of Reasoning: Assess whether the platform addresses only early-stage target identification or spans the full development pathway, including trial design, patient selection, and regulatory strategy.
• Deployment Model: Evaluate whether the platform is available through subscription access, enterprise licensing, or both, and whether it integrates with your existing research infrastructure and workflows.

Panna Sharma, Founder and CEO of Open-Medicine AI, emphasized the importance of this grounded approach. "We did not set out to build an AI company," Sharma stated. "We set out to get cancer drugs to patients faster and with less capital, and we built the tools we needed to do that because they did not exist".

Panna Sharma, Founder and CEO of Open-Medicine AI

"withZeta.ai is what came out of that work, and it is now running inside real programs with real regulatory and clinical consequences. Open-Medicine AI takes that architecture beyond rare cancers and beyond Lantern into new disease categories, new modalities, and the downstream parts of development where most of the time, cost, and failure live," Sharma explained.

Panna Sharma, Founder and CEO, Open-Medicine AI

What Does the Webinar Presentation Cover?

The September 23 webinar will detail four key areas of the Open-Medicine AI platform:

• Platform Architecture: The multi-agentic design behind Open-Medicine AI, how its co-scientists coordinate across the development pathway, and what distinguishes agents trained inside active clinical programs from general-purpose models.
• Development Roadmap: The next generation of co-scientists beyond rare cancers, the disease categories and therapeutic modalities being prioritized, and the downstream development challenges like trial design, patient selection, and regulatory work that the platform is being extended to address.
• Business Model: How Open-Medicine AI intends to generate revenue through subscription access and enterprise licensing, and how it works with biopharma companies, research institutions, service providers, and drug developers.
• Value Proposition: Where the platform is designed to remove time, cost, and failure from the path between a validated target and a treated patient, and how that impact is measured.

The webinar will be held Wednesday, September 23, 2026, at 12:30 PM Eastern time, with a 30 to 35 minute presentation followed by live question-and-answer discussion. A recording will be available afterward on Lantern Pharma's investor relations website.

How Does This Fit Into Broader Drug Discovery Advances?

The emergence of AI co-scientists trained in clinical environments reflects a maturation of AI drug discovery beyond early hype. The 2026 Lasker Awards recognized three scientists at Chugai Pharmaceuticals, Kunihiro Hattori, Takehisa Kitazawa, and Tomoyuki Igawa, for developing emicizumab, a bispecific antibody that transformed hemophilia A treatment. The drug reduced serious bleeding episodes by 87 percent in clinical trials and allowed patients to self-inject under the skin rather than receive intravenous infusions.

Similarly, the Basic Medical Research Award honored Emmanuel Mignot and Masashi Yanagisawa for their discovery of orexin, a neuropeptide critical to maintaining wakefulness. Their work led to the FDA approval of Takeda's Orzeyful in August 2026, the first approved treatment for narcolepsy type 1 designed to address the full spectrum of symptoms by directly targeting the loss of orexin signaling.

These breakthroughs demonstrate that the most impactful drug discoveries often come from researchers who follow unexpected findings rather than abandoning experiments when initial hypotheses fail. Yanagisawa's team was originally searching for molecules that activate orphan receptors in the brain, not investigating sleep biology. When they created mice lacking orexin, the expected changes in feeding and body weight did not materialize. Instead, the animals displayed behavior resembling human narcolepsy.

Open-Medicine AI's approach of embedding AI agents within active clinical programs reflects this same philosophy: building tools that work within the constraints and realities of actual drug development, rather than imposing external models onto the process. As the AI drug discovery market grows from 2.9 billion dollars to 13.8 billion dollars by 2033, platforms that can navigate the full development pathway while maintaining clinical rigor will likely become essential infrastructure for biopharma companies seeking to compress timelines and reduce failure rates.