Artificial intelligence (AI) is reshaping industries worldwide, but in pharmaceuticals, its adoption has been slower and more complex. Tightly regulated environments and concerns about patient safety mean that deploying AI is not as simple as integrating an off-the-shelf large language model (LLM).
For Dr. Chase Feiger, co-founder and CEO of Ostro, these constraints are not roadblocks but opportunities. His company partners with half of the world’s top 20 pharmaceutical firms, building AI solutions that adhere to strict compliance requirements while enabling more personalized interactions with healthcare professionals (HCPs) and patients.
“Regulations are there to protect patients and ensure accurate communication,” Feiger says in an interview with Drug and Device World. “They should be viewed as a foundation for innovation, not as barriers.”
Reframing the Challenge
While consumer tech companies can quickly adopt generative AI tools like ChatGPT,, Gemini, or Claude, pharma companies cannot take the same approach, at least when it comes to externally facing communications and materials. Every piece of promotional content that is shared with a healthcare provider or prospective patient must pass through medical-legal-regulatory affairs (MLR) review. It is also ultimately subject to prior US Food and Drug Administration (FDA) submission and review. Generative AI products, which produce new outputs with every user interaction, are simply not compatible with this process.
“All promotional content in pharma must go through a rigorous MLR review,” Feiger explains. “Generative AI creates new content with every interaction. Even without hallucinations, which are still a concern, those outputs are not pre-approved. They simply cannot be, by definition, if the technology you are using is generative. That alone makes direct use of generative AI impossible in this space.”
To overcome this, Ostro builds pharma-specific compliance directly into its AI-powered products. Drawing parallels with Veeva Systems, which added regulatory guardrails on top of Salesforce to create Veeva CRM, Feiger says Ostro’s platform ensures all content served to users is MLR-approved.
The challenge extends beyond MLR. Global data privacy and accountability regulations such as GDPR, CCPA, and the EU AI Act impose additional requirements for data privacy, minimization, and transparency. “Compliance is not optional,” Feiger says. “And for pharma brands that rely on building and maintaining the trust of the patient communities and healthcare providers that they serve, it’s not just a legal obligation – it’s an ethical and business necessity.”
Avoiding Strategic Missteps
The current hype around AI has led many organizations to launch pilot projects without clear objectives. Feiger cautions against this approach.
“Too often, companies focus on implementing AI because it’s trending, rather than asking what problem they are trying to solve,” he says. “The priority should be on identifying clear and specific clinical or business challenges and then determining how AI can help to address those challenges.”
He cites examples like reducing time to diagnosis, improving medication adherence, or accelerating patient and provider education efforts as areas where AI can add measurable value.
Aligning AI initiatives with clear outcomes also helps to secure buy-in across functions – compliance, medical, IT, and commercial operations – which is critical for success. “AI is not just about algorithms,” Feiger says. “It’s about aligning people and processes across the organization.”
Personalized Experiences for HCPs and Patients
One of AI’s most promising applications in pharma lies in tailoring information delivery to the needs of individual users.
“Not every doctor learns about a drug in the same way, and neither does every patient,” Feiger notes. An experienced oncologist, a first-year fellow, and a primary care physician all require different levels of detail and context when engaging with information about a complex therapy.
Ostro’s AI-powered engagement platform is designed to adapt to these differences, using data from previous interactions to adjust what content is presented. “For doctors, it means they get relevant information faster,” he explains. “For patients, it means receiving material they can understand and act on.”
Reducing Friction in AI Interactions
A common critique of AI tools is their reliance on user input quality. Poorly phrased prompts or insufficient context can lead to unsatisfactory experiences, which in turn discourage adoption.
Feiger says Ostro’s system addresses this by understanding the context in which users are interacting with the system, and by using this to interpret intent even when prompts are vague or contain errors. For example, when a user mistyped “Copaga” (the user intended to type “Copago” which means “copay” in Spanish, but misspelled the word), the AI correctly inferred the request – despite the misspelling and the fact that the brand’s materials did not include any Spanish language content – and surfaced appropriate and approved patient support resources.
“The goal is to minimize friction,” Feiger says. “Users should get the information they need, regardless of their familiarity with AI.”
Compliance as a Competitive Advantage
Rather than seeing pharma’s notoriously complex compliance regime as a burden, Ostro treats it as a differentiator. The company has achieved a 100% success rate in securing PRC (Promotional Review Committee) and MLR (Medical, Legal, and Regulatory) approval across its customers and can deploy its AI-powered solutions in under six weeks, even in the face of all of that complexity.
That’s because Ostro has invested from the earliest stages of the company in building products that are compliant and secure by design. “Regulatory compliance should be a moat, not a constraint,” says Dan Vorhaus, COO & General Counsel of Ostro. “It protects patients and providers, it protects our customers, and it builds trust in the brands that we support. For us, it’s been central to our product development from the beginning.”
As AI matures, Vorhaus and Feiger believe its role in pharma will grow, provided companies approach it thoughtfully. “The stakes are high in this industry because we’re dealing with people’s lives,” he says. “Getting AI right—ethically, responsibly, and compliantly—is not optional.”
