Publishers, both established and emerging, have experienced a turbulent few months with AI. There’s a clear love-hate relationship, whether it’s using AI or writing about it—it hurts. 

This concern was solidified this past week, as Microsoft’s news aggregation service MSN uploaded an AI-generated poll speculating on the cause of an Australian athlete’s death, besides a news piece by the Guardian. Obviously, this new development was not taken well by the publisher, and Microsoft was accused of damaging its “journalistic reputation.”

The readers too weren’t ready for it  – the “insights from AI’ turned the sensitive post into a engaging poll (see below). “This has to be the most pathetic, disgusting poll I’ve ever seen. The author should be ashamed,” read one of the comments from the user.  

Microsoft said that the generated polls for all news articles have been deactivated and it is investigating the cause of the inappropriate content. “A poll should not have appeared alongside an article of this nature, and we are taking steps to help prevent this kind of error from reoccurring in the future,” Microsoft general manager Kit Thambiratnam said.

The Pitfalls of Automating News

In MSN’s case, the team slowly shifted from employing over 800 editors human editors in 2018 to replacing them with what is believed to be AI-driven automation, laying off many in the process. The current number of editors—if any, at Microsoft is uncertain, but in 2020, the company laid off numerous journalists and editorial teams to replace them with AI.

Moreover, barely any human intervention in the process has resulted in them republishing articles including a claim that US President Joe Biden fell asleep during a moment of silence for Maui wildfire victims, a conspiracy theory about the Democratic Party orchestrating the latest COVID-19 surge, and an obituary for an NBA player describing him as “useless.” 

Additionally, Microsoft Start generated a travel guide suggesting visiting the Ottawa Food Bank in Ottawa, Canada, “on an empty stomach,” though the story was not entirely AI-generated but involved human review. 

Microsoft, previously, has also been accused of publishing junk content, such as stories about mermaids and Bigfoot sightings, after replacing human editors with automation. False articles from obscure websites were given greater visibility by Microsoft, and former editors suggest that without AI, such stories would not have been prominently featured. 

Same goes for NewsGPT, and humans too make mistakes. But, this is all changing. While MSN is trying to fix the recent concerns, others like Google News, have been trying hard to enhance the news reading experience for the better, where its algorithm is unintentionally biased toward more popular perspectives. 

At the same time, OpenAI is pushing hard for its product to be adopted across media outlets. It has already inked a pact with the Associated Press, one of the biggest news agencies in the US. Following that it announced over a $10 million partnership with the American Journalism Project (AJP)—said to explore ways to develop AI to support local news. 

Additionally, there are also murmurs of others like Google developing a tool, internally known as Genesis, which is designed to generate news content from provided information, potentially assisting journalists by automating tasks. Google envisions this tool as a responsible technology that could help the publishing industry navigate the challenges of generative AI. 

The Birth of New Age Media 

On the bright side, AI is finding its place in journalism worldwide. A recent global survey which collected insights from over 105 newsrooms in 46 different countries revealed that about 90% of newsrooms are already using AI in news production. They are using AI for transcription, text extraction from images, writing articles, translating content, proofreading, generating headlines,  or making it more engaging (like the one you saw above) – “insights from AI”

The survey conducted by a think tank at the London School of Economics, supported by the Google News Initiative further revealed that 80% of it is being used in news distribution which utilises AI-driven search engine optimization, and 75% in news gathering which includes automated transcription, translation, text extraction from images, and web scraping. 

However, over 60% of respondents expressed concerns about the ethical implications of AI integration in journalism—indicating that newsrooms generally view human intervention as vital for mitigating potential issues such as bias and inaccuracy caused by AI systems. 

The New York Times, The Washington Post and The Wall Street Journal amongst others are experimenting with AI. Gannett, the largest newspaper publisher in the US is also using AI to write short stories for print based on stories written by human writers. At AIM, we have also been experimenting with integrating a few AI capabilities, mostly for internal use, alongside enhancing productivity, content and engagement. These are surely exciting times. 

All in all, we are on the brink of a new media era: the fusion of AI with human creativity, and it’s nothing short of surreal. It’s time we welcomed this union with open arms, just as we’ve embraced SEO, social media, and other information-sharing platforms.