Ohio State University journalism professor Nicole Kraft asked a question at an SPJ 2026 panel, “AI in the Newsroom and Classroom: Practical Uses, Ethical Guardrails,” that is on the minds of many journalists: “Who does not want to use AI?”
Most of the audience raised their hands.
Kraft then asked who used AI lightly, such as for grammar and spell-checking.
Fewer hands went up.
Finally, she asked who uses AI for everything.
Only one person raised their hand.
Audience members point their thumbs down when asked if it is ethical to ask AI to pull quotes from an interview transcript. Photo by Juan Carlos Ramirez Delgadillo.
The exchange showed the uncertainty journalists face as AI becomes increasingly integrated into newsrooms.
Panelists discussed the ethical boundaries of AI in journalism, including when reporters can use the technology to assist with research and reporting and when its use begins to replace human judgment.
They included Kevin Smith, director of Ohio University’s Kiplinger Program in Public Affairs Journalism; Chad Painter, associate professor and chair of the University of Dayton’s Department of Communication; and Peter Baniak, the AI Cross Faculty Fellow and visiting professor at the University of Kentucky.
They discussed the ethical boundaries of AI in journalism, including when reporters can use the technology to assist with research and reporting and when its use begins to replace human judgment.
AI is an extremely powerful tool, but it comes with many risks, said Baniak. A split in opinion comes from how it is being used. AI can help understaffed newsrooms monitor public meetings and analyze large amounts of information, they said, and useful for data collection and analyzing, but people also use it for content creation.
One of the panel’s central principles was what Baniak described as the “human in command” imperative, which makes people responsible for verifying what AI produces.
“No matter how you use artificial intelligence, no matter what you are using it for, ultimately the human is in charge, and the human is responsible for the output,” Baniak said.
For Smith, AI developments move extremely fast. He compared it to attempting to track down a bullet train. What people learned last year about AI has exponentially blown up as more new ways to use AI emerged this year, making it a challenge to keep up, Smith said.
“We just have to make the decision as journalists on what we’re willing to live with ethically or legally as this thing continues to go,” Smith said.
Panelists and audience members discussed how they use AI in their work. Some said they use it to transcribe interviews, while others use it for headlines and source finding.
Smith said his newsroom’s content management system had a built-in AI headline writer. While the headlines were poorly written, it made him realize that they could do better, Smith said.
“It helped us be more human because we realized how bad it was,” he said.
Educating people on AI comes in levels, Painter said. People need to understand AI systems and interpret the outputs. After establishing some AI fluency, they need to be able to use the tools.
The last part is professional accountability, which Painter said often gets short-changed, but it is actually the most important. “It’s not just ‘I know what it is and I know how to use it’ but rather ‘I know when I should and when I shouldn’t,’” he said.
The panel included an activity where the audience listened to different ethical scenarios and voted on whether the use of AI described was ethical, not ethical or in need of more consideration.
Scenarios included using AI to generate interview questions, identifying possible angles from long documents and quoting extraction from interview transcripts.
The scenarios discussion led attendees to believe that humans have something AI doesn’t: creativity and intuition.
“AI will never come up with something as good as a headless body found in topless bar,” Painter said.