How an Industry-Backed Researcher Uses AI to Challenge Pollutant Dangers

An industry-backed researcher who has built a career sowing doubt about the dangers of pollutants is now leveraging artificial intelligence (AI) to amplify his perspective. Louis Anthony “Tony” Cox Jr, a Denver-based risk analyst and former Trump adviser, is developing an AI application to scrutinize academic research for what he perceives as the false conflation of correlation with causation.

Cox, who has previously claimed there is no proof that cleaning air saves lives, is working on an AI tool to filter out what he calls “propaganda” from epidemiological research. Emails obtained via Freedom of Information Act requests by the Energy and Policy Institute and reviewed by the Guardian reveal Cox's intentions to perform “critical thinking at scale.”

Cox has a history of questioning the robustness of research linking exposure to chemical compounds with health risks. He has worked on behalf of polluting interests, including cigarette manufacturer Philip Morris USA and the American Petroleum Institute, a fossil fuel lobbying group that has even been allowed to “copy edit” his findings. Despite these collaborations, Cox notes that he has also secured public research funding.

The tobacco and oil industries have a well-documented history of weaponizing scientific uncertainty, according to experts. These tactics have been linked to the Trump administration’s deregulatory efforts. For instance, the president’s May “gold standard” science order gave his appointees the power to “correct scientific information” and “discipline” those who deviated from the administration’s views, sparking outrage among scientists.

Cox has received funding from the American Chemistry Council (ACC), the nation’s largest chemical industry advocacy group, to develop the new AI reviewer. The ACC includes oil and chemical giants such as Exxon and DuPont among its members. This sponsorship raises questions about the project’s impartiality.

When asked about these concerns, Kelly Montes de Oca, a spokesperson for the ACC, stated that the research aims to “support scientific understanding and analysis of chemical exposure and human health, enhance transparency and reproducibility, advance the safety of chemical products and processes, and inform science-based global regulatory approaches.”

Cox told the Guardian that his AI assistant is designed to help users understand the objective implications of data without distortions from human biases. He added that his work aims to use sound technical methods to pursue scientific truth. However, critics argue that the ACC is not a neutral force and seeks to minimize regulatory burdens on its members.

In February 2023, Cox engaged in a conversation with the AI assistant ChatGPT, asking it a series of questions about PM2.5, a fine particulate matter linked to various health issues. Despite the bot’s response noting strong scientific evidence of the link between PM2.5 and lung cancer, Cox continued to probe for potential confounding factors. He also requested sonnets and posed riddles during the conversation, which was later sent to his University of Colorado email address.

Cox’s AI tool is currently being tested on submissions to academic journals, including Risk Analysis, which he edits. He maintains that the tool is designed to evaluate research submissions before they undergo peer review, ensuring a more objective analysis. However, critics remain skeptical of the project’s funding and potential biases.

Chris Frey, associate dean for research and infrastructure at North Carolina State University’s College of Engineering, who chaired the Environmental Protection Agency’s clean air scientific advisory committee from 2012 to 2015, stated, “The ACC lacks impartiality as they aim to minimize regulatory burdens on their members.”

While Cox asserts that his tool has no agenda and is transparent about its funding, the skepticism surrounding the project highlights the ongoing debate over the role of industry funding in scientific research.