“Trust” Key for Agentic AI in News

“In the media and broadcast industry, we are at the very early stages of the whole agentic AI journey in some ways,” said Martins Magone, CTO, VESET. 

Moderator Rob Ambrose, Co-Founder, Caretta Research, noted that governance of AI “perhaps is more crucial in the news than in any other part of the media. Trust is such an important part of news.” 

“Reputation and trust are absolutely paramount,” agreed Mahesh Ramachandran, CTO, Reuters. “Everything we do [at Reuters] is underpinned by the five trust principles of that. First of all, no AI use case can even go into a development stage unless it is approved by the AI governance committee. 

“Things don't go far if it doesn't pass muster,” he added. “Secondly, every AI tool, even if a human in the loop is there, needs to pass a very high bar for the journalist. [It needs to be] the same quality that we expect journalists to do.” 

On a transition, he said: “You bring the journalists into the process…  It’s a three-stage [process]. You take away the fear, define the purpose, and pull [them] into the business. That's what actually makes it successful.” 

Ryan Currier, Vice President, Product Management, Amagi, reported that he is seeing “strong desire” to reduce or get rid of “laborious” tasks, “so there could be a greater focus on journalism, creativity, unlocking distribution, and maybe audiences that they aren't currently reaching because they didn't have time to focus on it.” 

“I think it's definitely happening,” Ramachandran added, speaking about the process of bringing agentic AI to news. “If I can be a bit provocative, I'll say that the shift to agentic is going to make a lot of technologies obsolete.”