Adobe made waves at their Summit 2025 by introducing AI agents designed to improve customer experiences. These smart tools analyze data to predict what people might buy or need, helping companies serve clients better. Meanwhile, OpenAI shared findings about their Deep Research system, which can browse thousands of web pages without getting tired. While it outperforms humans in marathon research tasks, it still makes mistakes nearly 50% of the time, showing room for improvement.

Researchers worldwide are preparing for a major AI science workshop in Singapore later this month. The event will tackle tough questions like how to help AI systems share knowledge across biology, chemistry, and physics.Experts will also discuss ways to ensure AI discoveries can be trusted and repeated in real labs.

In education news, Clarivate expanded its academic platform with AI agents that help students write papers and find reliable sources faster. These tools act like digital librarians, sorting through millions of articles to suggest the best references. Early tests show these agents could cut research time by half for complex projects.

Scientists are excited but cautious about these developments. While agentic AI shows promise for breakthroughs like new medicines or climate solutions, challenges remain. Teaching AI to ask good questions and work safely with human researchers will be key focus areas in coming years.