The AI paradox

Since generative AI hit the headlines in 2022, few industries have felt the impact of the hype and hysteria that accompanied the technology more than news. At CMT we documented the early reactions of Australian newsrooms in our 2023 report, Gen AI and Journalism, finding deep trepidation about the risks of AI mixed with cautious optimism about the potential opportunities. Last week, we released our second report. Eighteen months down the road, there’s more nuance and hard-headed realism about the potential impact of generative AI. The report distils insights from interviews with 19 news editors and product managers from 14 newsrooms, as well as a day-long workshop that brought newsroom leaders together with tech companies and industry researchers. We devote this week’s newsletter to some of the key issues highlighted in the report.

First, Monica looks at how the news industry is dealing with the need to take a collaborative approach as it recognises the common needs, and in some ways, a common enemy, brought by AI disruption. For our participants, the most pressing concern about AI is not so much the risk it presents to journalistic integrity or the potential to replace journalists, though these are also front of mind. Rather, it is concern over the loss of editorial control and potential drop in traffic to news websites as AI companies increasingly assume control over news distribution. This problem has been brought into stark relief by the Productivity Commission’s report, released last week, which has called for feedback on a proposed text and data-mining (TDM) exception to copyright. News organisations and creative industries are incensed.

Providing context on this issue, Tamara looks at relevant developments in copyright around the globe, focusing on the raft of cases currently before the courts that could seriously impact AI companies if judgements go against them. Finally, with AI adoption notably slower in Australia than elsewhere, I look at whether Australian newsrooms are falling behind the game. While they are certainly cognisant of the potential for AI to improve workflow efficiencies and save costs, many newsrooms have found that, given the limitations of AI tools and the need for robust oversight, the benefits still don’t outweigh the costs.

Putting these together, our research reveals the paradox that newsrooms are confronting: AI is both a means of eking a bit more productivity out of perpetually declining revenue, and a growing threat to that same revenue.

Read the newsletter