Each issue rounds up the month's most notable AI research touching chemistry, chemical engineering, and science education, drawn from our research dossiers. Browse by year below, newest first, and click any issue to read it in full.
July was a quiet month for AI hype and a loud month for AI homework checking. A drug-screening pipeline got caught picking candidates no better than random guesses, and a robot lab needed a second AI just to supervise the first.
Read the issue June 2026June was the month AI had to account for geometry, physics, variable water, and students who were actively trying to make it fail. A new language model gave transition metal complexes a 3D view.
Read the issue May 2026May was the month AI had to leave the demo and touch something real: a small diagnostic device, a robotic materials loop, and new rules for engineering education.
Read the issue April 2026April's AI research had a practical mood. Models had to keep data private, respect thermodynamics, survive missing sensors, and avoid unsafe chemical routes.
Read the issue March 2026March put AI in chemistry through four increasingly unglamorous tests: a polymer that exists, shared public infrastructure, critical metals recovered from waste, and a retraction.
Read the issue February 2026February's AI research had a repeated request: show your work and stay inside the laws of physics. Catalyst models extracted a chemical rule, and process agents checked their own calculations.
Read the issue January 2026January was the month AI science stopped pretending the model was the whole system. Governments funded shared laboratories, and a chemistry exam found ChatGPT handled the word “NOT” better than the students.
Read the issueNovember's most useful finding was a negative one.
Read the issue October 2025October was about making things that work in a real building. A screening framework found iron-copper and nickel-molybdenum catalysts that match ruthenium without the ruthenium price.
Read the issue September 2025September delivered the industrial numbers everyone had been asking for. CATL validated 200 battery formulations a day. Tesla cut a test cycle from 70 hours to 6.
Read the issue August 2025August was a peer-reviewed month, which makes a pleasant change. IBM built a molecular generator fast enough to produce ten billion structures and then honestly reported what happens when you do.
Read the issue July 2025July's best idea was to stop waiting. A self-driving lab that used to idle up to 60 minutes per experiment started measuring during the transitions instead and got ten times the data.
Read the issue June 2025June was the month the tooling got given away. Microsoft released a neural functional and the dataset behind it. MIT CSAIL and Recursion open-sourced a binding affinity predictor under the MIT license.
Read the issue May 2025May produced the first peer-reviewed head-to-head between language models and expert chemists, and the models won on points. They also learned nothing about when to hesitate.
Read the issue April 2025April was the month the field stopped adding parameters and started adding rules. MIT modeled all of hydrogen combustion with 344 of them. A 7-billion parameter chemistry model beat the general-purpose giants by knowing one thing well.
Read the issue March 2025March was a month of AI doing less with more. A battery startup replaced an estimated 50 years of physics simulation with under 24 hours of generative modeling. A chemistry model beat the standard benchmark using roughly a tenth of the training data.
Read the issue February 2025February was the month AI stopped answering questions and started holding debates. Google built a system where specialized agents argue over hypotheses in ranked tournaments until one survives.
Read the issue January 2025January was the month chemical AI stopped grading itself on a curve. Researchers spent it building benchmarks designed to catch their own tools cheating, and the tools obliged.
Read the issue