Regulatory, medical writing and pharmacovigilance

Takeda is deploying Weave Bio's AI regulatory-writing platform across 14 programs. Use started with IND drafting and has grown to health-authority question responses.

Partners: Weave Bio

Novartis says generative AI can create well-structured first drafts of clinical and regulatory documents from approved sources, which Novartis experts then review and refine.

Daiichi Sankyo says that, going forward, it is renewing its safety database into a next-generation database that incorporates automation by AI, in order to make global pharmacovigilance operations more efficient and to strengthen analysis.

Pfizer scientists published a generative AI workflow that drafts pharmacokinetic sections of clinical study reports and scored its output against manually written reports.

J&J's CIO said AI now handles much of the preparation of clinical trial reports for regulators, cutting a job that took hundreds of hours to minutes.

BeOne says it leverages assistive AI and digital tools to analyze safety data more efficiently and to speed the identification of potential risks. The statement sits in the report's description of its in-house global patient-safety and pharmacovigilance work.

In the same R&D Day question session, Astellas said each team is already using AI to generate clinical study reports. The same answer says AI translation has sped translation of clinical-trial documents, with human confirmation still required.

Shionogi said a generative-AI solution for drafting regulatory documents in drug development was built by combining its medical writers' knowledge and its data scientists' AI design work with Hitachi's generative-AI know-how. A proof of concept at the company confirmed about a 50% reduction in clinical study report preparation time and about a 20% reduction in protocol preparation time. The same release says Shionogi licensed the solution to Hitachi, which began offering it to pharmaceutical and healthcare companies in Japan in February 2026.

Partners: Hitachi

Sanofi says it uses machine learning and AI across R&D, including for writing clinical study reports. No tool or metrics are named on the page.

Sobi's capital markets presentation lists generative-AI authoring of clinical study reports and Module 2 on a map of technology it describes as enabling faster development-cycle time.

A custom LLM application that drafts informed consent forms from a trial protocol and template, with human review. Vertex data scientists report it has supported more than ten trials.

Vertex's CEO said AI is helping streamline work such as preparing regulatory filings, which usually take six to nine months. Reported secondhand from a public event.

An in-house generative AI documentation platform built on Claude via Amazon Bedrock with MongoDB Atlas for drafting regulatory documents such as clinical study reports. Anthropic reports large time savings.

Partners: Anthropic, AWS, MongoDB

Merck built an internal generative AI tool that drafts clinical study reports for human review, more than halving first-draft effort, and has submitted CSRs produced this way.

A generative AI tool that helps write Sanofi's annual product quality reviews (PQRs). In June 2025 Sanofi reported about 3,500 PQRs a year and a target to cut creation time by 70 percent. An Aug 2024 Outsourced Pharma interview with Sanofi reported PQR writing about eightfold faster, a reported result and not a target.

During 2024, Viatris continued exploring emerging technologies, including artificial intelligence, to potentially enhance product safety evaluation, communication and risk mitigation.

Sanofi's pharmacovigilance team uses AI and OCR on the IQVIA Vigilance Platform to automate intake of the more than 700,000 adverse reactions it receives each year.

Partners: IQVIA

Takeda's medical writing team built a generative AI prototype to speed compilation of regulatory dossiers by automating analysis of clinical study data. It was being tested on a CSR efficacy section.

AstraZeneca built a knowledge graph combining public adverse-event and exposure data to speed COVID-19 vaccine safety analyses and periodic reporting.