Google is facing criticism regarding the departure of a researcher who raised concerns about potential biases within artificial intelligence. A recent report by Reuters indicates that other employees engaged in similar research at the company have been requested to maintain a “positive outlook” and are now subject to more thorough scrutiny when their work addresses “sensitive subjects.”

According to Reuters, referencing conversations with researchers and internal documentation, Google has introduced new oversight measures over the past year. These include an additional review process for publications covering specific areas and a noticeable increase in executive involvement during the later phases of research projects.

This situation closely mirrors the events surrounding Dr. Timnit Gebru, an AI researcher at Google. Her departure from the company appeared to be involuntary, occurring amidst unclear circumstances and disagreements with leadership concerning her team’s work. (I have reached out to both Gebru and Google for statements regarding this matter.)

An internal webpage reviewed by Reuters identifies several “sensitive” topics, including: “the oil industry, China, Iran, Israel, COVID-19, home security, insurance, location data, religion, self-driving vehicles, telecommunications, and systems used for recommending or personalizing online content.”

While it’s apparent that many of these areas are inherently sensitive, instructing researchers to exercise caution when exploring them may seem unnecessary given the established presence of ethics committees, peer review processes, and other standard research controls. A researcher who spoke with Reuters expressed concern that this level of intervention from Google could lead to “a significant censorship issue.”

This development also highlights the broader challenge of conducting crucial research within a corporate environment where publication may be influenced by the company’s interests. While large, privately funded research institutions have a long history, companies such as Facebook, Google, Apple, and Microsoft wield substantial influence over fields like AI and have a vested interest in promoting the benefits of their profitable technologies while minimizing potential criticism.