The quasi governmental body that advises the U.S. government on science is fighting over a climate chapter in a handbook that federal judges use to weigh expert testimony.
The National Academies of Science, the quasi-governmental body that has advised the U.S. government on scientific questions since the Civil War, is in the middle of a public fight over a single chapter in a reference manual that federal judges use to weigh scientific evidence.
The dispute, reported this week by Ars Technica, centers on the climate chapter of the fourth edition of the Reference Manual on Scientific Evidence, a handbook prepared for the Federal Judicial Center to help judges handle expert testimony. Republican state attorneys general objected to the chapter on grounds that some authors have ties to climate litigation, and that treating human-caused climate change as established science is not impartial. The Federal Judicial Center pulled the chapter in February 2026 after receiving a letter from the AGs. The Academies kept its copy on the NAS website and issued a two-sentence response in March, saying it used the same procedures as every other chapter in the manual.
This is the kind of pressure the institution was built to absorb. The Academies' century-old model is straightforward: convene top scientific talent, produce the most rigorous reports available, stay above political fights, and let the resulting prestige do the shielding. The model has held through politically sensitive reports on tobacco, acid rain, climate change, and vaccines, and the federal agencies that fund the institution have generally continued to fund it.
What is different now is the second object in play. The Academies is also preparing a separate expert report on attributing individual weather events to human-driven climate change, a field of science that has direct bearing on active liability litigation against fossil fuel companies. The same Ars Technica analysis notes that industry attorneys are concerned the attribution report could support findings of liability in those cases. That makes the climate chapter in the judges' manual a related but distinct front in a larger fight about which evidence reaches a courtroom, and on what terms.
The legitimate critique embedded in the pressure campaign is real. A reference manual written for judges should be defensible against the appearance of bias, and authors with active roles in litigation have an obvious conflict to disclose. The harder question is whether the Academies' response protects the institution or narrows it. It can defend the chapter on process grounds, narrow the scope, publish a public explanation, separate litigation expertise from bias, or accept the revisions. Each option changes what the science referee looks like to the next generation of judges, jurors, and policymakers who will live with whichever version emerges.
Watch the next move at the Federal Judicial Center. Distribution of the manual was paused after the AGs' letter; whether it resumes unchanged, in revised form, or under a different arrangement will tell the reader how much insulation the Academies still has, and how much it is choosing to spend.