An initiative focused on partnering with retailers to decrease the influence of tobacco, alcohol and lottery marketing is showing promise and could present a more effective way to protect public health throughout the country, according to researchers at The Ohio State University.

The program, called Retailers Creating a Responsible Environment (Retailers CARE), launched in Montgomery County, Ohio, and may be the first effort in the nation to incentivize retailers to voluntarily reduce promotion of tobacco, alcohol and lottery products.

Following the earlier deployment of MMAI Gym for Science for large language model (LLM) post-training, Insilico Medicine has expanded its generative AI capabilities through a strategic collaboration with Liquid AI, yielding a 2.6B parameter model optimized for retrosynthesis and a high-capacity 24B parameter model, both available for on-premise deployment and via the Microsoft Marketplace. To enable enterprise-grade automation, Insilico has unified its entire Pharma AI suite via the Model Context Protocol (MCP), allowing researchers to execute complex workflows directly within developer environments like Claude Code and Cursor. Concurrently, the company launched PandaClaw, an autonomous AI agent for multi-omics target discovery, alongside critical upgrades to its core suite: dynamic multibody simulation scaling in MDFlow, an expanded kinase profiling panel, advanced ionization states in generative chemistry, and enhanced MD simulations for metalloproteins.

A Dartmouth study finds that AI may cost doctors time when corresponding with patients by making them spend more time editing AI-drafted responses than it would take to write them. The researchers analyzed 146,000 conversations between patients and their primary care physicians from the online portal of a large healthcare system and found that automated responses frequently introduced errors and extraneous details. But the study provides a framework for generating "good" AI responses that align with how clinicians communicate and shows that adapting AI to individual physicians can reduce editing by 26%. The study was presented at the 64th Annual Meeting of the Association for Computational Linguistics.