Table of Contents 64q43
A FDA (acronym in English for Food and Drug istration), the United States Food and Drug istration, is in discussions with the OpenAI the use of artificial intelligence to speed up the process of evaluating new drugs. The idea is to apply language models that work like the Chat GPT, but aimed at automating parts of the analysis of documents and data related to drug evaluation, reducing the time required in some final stages of the review. Understand: y3s5r
Negotiation between OpenAI and FDA 1k6cr
In recent weeks, representatives of the OpenAI and FDA participated in several strategic meetings to discuss the use of artificial intelligence (AI) in the agency's internal processes, according to the portal Wired.
One of the main focuses of these meetings has been the project called cderGPT, which refers to the Center for Drug Evaluation and Research (CDER), responsible for approving prescription and over-the-counter drugs in the country. The proposal involves the possibility of using AI to optimize and speed up the analysis of data and documents related to the development and approval of new drugs.
In addition to the technical team of OpenAI, representatives linked to the so-called participated in the conversations Department of Government Efficiency, initiative associated with Elon Musk, and Jeremy Walsh, newly appointed first Chief AI Officer at FDA. Although no formal agreement has been signed yet, the negotiations signal a move by FDA for the modernization of its regulatory processes, especially with regard to scientific evaluation and the management of large volumes of data, a topic already discussed previously by Marty Macary, FDA:
We at the FDA now have to ask big questions we’ve never asked before. Why does it take more than 10 years for a new drug to come to market? Why haven’t we modernized with AI and other technologies? We just completed our first AI-assisted scientific review of a product, and this is just the beginning.
FDA Commissioner Marty Makary on his X profile.
This interest in AI-based solutions follows previous initiatives by OpenAI, such as the launch of ChatGPT Gov, an adapted version of its chatbot for government use, with a focus on meeting data security and privacy requirements. The use of tools such as Research GPT, which was also reportedly mentioned in the discussions, can faster and more accurate analyses of clinical research and scientific literature.
Furthermore, OpenAI also works towards obtaining program certifications FedRAMP (Federal Risk Management and Authorization Program), which regulates the use of cloud services by federal agencies. Obtaining certifications at the moderate and high levels would allow the OpenAI store and process sensitive government information securely, expanding the possibilities for collaboration with public institutions such as FDA.
cderGPT Project 6l6nm
The project cderGPT, currently under discussion between the OpenAI and FDA, aims to explore how artificial intelligence can make the complex process of evaluating and approving drugs more efficient. The name refers to CDER, the agency's department that regulates over-the-counter and prescription drugs. The proposal is that cderGPT act as a tool for human reviewers, speeding up bureaucratic and analytical tasks, without compromising the strict safety and efficacy criteria required in the healthcare sector.
In practice, the cderGPT could automate activities such as checking the completeness of documents sent by pharmaceutical companies, initial screening of clinical data and even in the analysis of comparative efficacy between treatments. This does not mean replacing agency professionals, but rather offering technical so that reviewers can focus on more strategic and complex decisions. In this way, it is expected to significantly reduce the response time in certain stages of the process, especially in cases of drugs aimed at serious conditions or with few therapeutic options available on the market.
A FDA already recognizes the potential of artificial intelligence and, since December 2023, has been directly funding research on the topic. The agency offered a grant for the development of large language models (LLMs) aimed at internal use. The researcher involved in this program is primarily focused on applying these models in areas such as precision medicine, drug development, and regulatory science. These studies seek to prepare the agency to integrate AI safely and efficiently into its processes, including before submitting drugs for formal review.
Despite the enthusiasm, it is worth ing that AI would only be used in the final stages of the review process. Most candidate drugs still fail in the early clinical phases, well before reaching the FDA. Even so, tools such as cderGPT can optimize the documentary and analytical part of the process, complementing existing acceleration mechanisms.
The FDA's Role in Drug Control 573m5o
In the United States, Food and Drug istration (FDA) is the federal agency responsible for the regulation and control of medicines, food, cosmetics, biological products and medical devices. In the specific case of medicines, the FDA operates from the preclinical phase, monitoring laboratory and animal testing, to the clinical stages in humans, which occur in three main phases. Only after the successful completion of these stages can companies submit a formal approval request (NDA – New Drug Application). From there, the review process by FDA It can take months or even more than a year, depending on the complexity of the product and the regulatory category it falls into.
the performance of FDA It also includes factory inspections, post-market safety monitoring (pharmacovigilance), and issuing alerts or even withdrawing products from the market if necessary. The agency is known for its technical rigor and its requirement for robust evidence of efficacy and safety, which helps ensure the confidence of healthcare professionals and the American public in approved products.
The agency already has specific programs to speed up the approval of drugs aimed at serious diseases or those with no treatment options, such as fast track, which facilitates communication between the FDA and companies to speed up drug development; breakthrough therapy, intended for treatments that show promising results in early stages; and priority review reduces the time for final analysis from ten to six months in cases considered to be a priority.
In Brazil, the person who performs a similar function is National Health Surveillance Agency (ANVISA). Just like the FDA, ANVISA is responsible for the approval of new medicines, analysis of clinical dossiers, health inspections and post-marketing surveillance.
What do you think of this advancement in the use of AI in drug assessments? Tell us about it Comment!
See also:
Sources: TechCrunch e Wired.
Text revised by Felipe Faustino in 08 / 05 / 2025
