AI in Action | May 2025
Artificial intelligence (AI) has the potential to help us solve the greatest challenges of our time from health care, agriculture, and education to transportation, energy, and national security.
When guided by ethics, transparency, and fairness, responsible AI can spark creativity and help usher in a more prosperous and sustainable future.
See how AI is improving our lives, growing our economy, and keeping us safe.
AI in Action
MIT is making AI-generated code more accurate in any language
Researchers at MIT have developed a new approach to help LLMs write cleaner code. This method automatically guides AI to generate text that follows the rules of a specific programming language and avoids errors that can lead to computer crashes. The tool works by assigning probabilities to different code outputs and steering the AI toward the ones most likely to be valid. It can also stop early when a promising path isn’t panning out, saving time and computing power. This smarter, more efficient approach promises to make AI-powered coding tools faster, safer, and more helpful for developers.
Learn more.
Etsy marries AI and human expertise to create curated product collections
Etsy has introduced a new browseable search feature within its app, combining AI with human expertise to create curated product collections. Initially, experts select around 50 items based on trends, aesthetics, and occasions. Machine learning then expands these to approximately 1,000 listings, with LLMs ensuring each collection is cohesive, diverse, and meets quality standards. This initiative builds upon Etsy’s previous AI-driven holiday gift recommendation tool, reflecting the company’s ongoing efforts to enhance personalization and user engagement through technology.
Learn more.
Isomorphic Labs is utilizing AI to redefine drug discovery
Isomorphic Labs, a drug discovery company, is using an AI-first approach to rethink how we understand biology and develop new medicines. By modeling biological systems at the molecular level, researchers can simulate how potential drugs interact with targets in the body without having to step foot in a lab. This enables scientists to predict outcomes with remarkable accuracy, accelerating the timeline from discovery to treatment. With fewer barriers and faster testing, Isomorphic is opening doors to therapies for diseases once thought untreatable.
Learn more.
Doctors told him he was going to die. Then AI saved his life.
AI, automation helped these cities transform operations
TSA tests out AI to train, assist airport screening officers
AI Spotlight
OpenAI has updated its Preparedness Framework to better identify and mitigate severe risks from advanced AI, focusing on threats that are plausible, measurable, and potentially irreversible. The framework categorizes high-risk capabilities—like biosecurity threats and autonomous AI behavior—into tracked and emerging areas, guiding responsible development. By enhancing testing, governance, and transparency, OpenAI sets a precedent for safe AI deployment. This initiative is crucial for protecting society from misuse while fostering trust and accountability in AI technologies.
Learn more.
About AI for America
AI for America is an initiative to educate the public about how AI is being used to improve lives, grow our economy, and keep us safe. The initiative combines coalition building, advocacy, social media, and traditional media to demonstrate the immense economic and societal benefits of AI. AI for America is funded by TechNet, the national, bipartisan network of innovation economy CEOs and senior executives. TechNet advances public policies at the federal and state level that make the United States the world leader in innovation. Its diverse membership includes dynamic American businesses ranging from startups to the most iconic companies on the planet. For more information, go to technet.org and AI4America.com.