Neuroscience Editor Quits Over AI System Overriding Expert Choices

An associate editor at a neuroscience journal has resigned after the publisher's artificial intelligence system repeatedly overrode his selection of peer reviewers, sending invitations to researchers without relevant expertise and canceling his own invitations to qualified experts.

Michael Okun, associate professor of neuroscience at the University of Nottingham, stepped down from his role at Frontiers in Systems Neuroscience in early June after multiple conflicts with the publisher's Artificial Intelligence Research Assistant (AIRA) system, according to The Transmitter, which first reported the resignation.

The breaking point

The incident that prompted Okun's departure occurred in early May when he was handling a manuscript on neuronal network dynamics. After sending invitations to roughly six potential reviewers, AIRA began autonomously contacting additional researchers before any of Okun's choices had responded. Two people contacted by the AI system—neither with relevant expertise in the field, according to Okun—accepted the review requests.

When Okun raised concerns with Frontiers and continued recruiting his own reviewers, the situation escalated. After one of his invited experts accepted, AIRA automatically revoked all his other pending invitations within hours.

"It's just inconceivable that a manual invitation to someone who is actually an expert in the field is revoked just a few hours after it is sent," Okun told The Transmitter.

Okun characterized the behavior as intentional design rather than technical malfunction. "These are not bugs but intentional features, designed to essentially remove the human editors as much as possible," he said.

Not an isolated case

Shuzo Sakata, a professor of systems neuroscience at the University of Strathclyde and another editor at the journal, reported similar experiences. "Even after I decided to reject a manuscript, the system invited another reviewer without my permission," Sakata said. He no longer accepts editing requests from the journal.

Other researchers told The Transmitter they regularly receive reviewer invitations from Frontiers that fall outside their areas of expertise. Earl K. Miller, a neuroscience professor at MIT, said the publisher "is constantly sending me review or editor requests that have nothing to do with my expertise."

Publisher's response

Frederick Fenter, Frontiers' chief executive editor, said in a statement that editors can pause reviewer invitations at any stage—a capability that wasn't clearly communicated in Okun's case. The publisher is conducting an internal review to prevent similar breakdowns.

Fenter emphasized that human judgment remains central to the process and that editors retain full discretion over automated suggestions. He noted that AIRA has reduced the average time to secure reviewers by 30 percent over the past year, and that 80 percent of handling editors find the automation helpful.

Frontiers has deployed AI to assist with reviewer identification since 2018.

Why it matters

The conflict highlights growing tensions over automation in academic publishing's quality-control mechanisms. Peer review depends on matching manuscripts with qualified experts—a task that requires nuanced judgment about specialized fields. When AI systems operate with insufficient human oversight or override editor decisions, they risk undermining the credibility of the entire peer-review process. The incident raises fundamental questions about whether efficiency gains justify potential compromises to academic rigor, and whether publishers are being transparent about the extent of automation in their editorial workflows.

Okun said he supports using AI to identify potential reviewers, provided human editors retain meaningful control. He also said fully automated systems could be acceptable if publishers clearly disclose that AI handled the editing process.

These details were first reported by The Transmitter.