Our Hometown Blog

Introduction

A report from the London School of Economics, “AI & Journalism: New powers, new responsibilities” provides a comprehensive look at how news organizations around the world are adopting artificial intelligence technologies. The Journalism AI Report was compiled by Charlie Beckett, founding director of Polis, and surveyed over 70 media companies in 32 countries about their use and views on AI. It found that while AI is already deeply embedded in many newsrooms, its future impact remains uncertain. News organizations are trying AI for tasks like automated content tagging, production, and distribution, but they do not see it yet as transformational.

Definitions of AI Vary

The report found that newsrooms tend to define AI based on how they use it, not just on technical aspects. For example, some focus on machine learning’s pattern recognition abilities, while others see AI as performing human-like cognitive tasks. Many stressed that AI is there to augment human capabilities. The report advises newsrooms to settle on a definition of AI to help shape strategy and communication about it internally. Of the respondents, 25% said they did not have a formal AI definition yet. Developing a common understanding of what AI means for your newsroom is an important first step.

Current Uses Improve Efficiency and Relevance

When asked how their organization currently uses AI, the top responses related to newsgathering (45% of respondents), news production (68%), and distribution (51%). Applications ranged from automated content tagging and personalization engines to AI-generated text, image and video content. The main motives given for using AI were improving efficiency (68% cited this), delivering more relevant content (45%), and enhancing business performance (18%). Media companies are deploying AI widely, but mostly to supplement routine journalism tasks right now.

Adoption Faces Challenges

Despite active experimentation with AI, the report identified serious barriers to its adoption for many newsrooms. The top challenges cited were lack of financial resources (27%), gaps in skills and knowledge (24%), and organizational culture resistant to changing tools (24%). Management issues and lack of strategic planning around AI were also cited as key obstacles. For example, 37% of respondents said they have an AI strategy so far. So while journalists may be keen to test AI, they face real barriers in metrics, training, and leadership support.

Ethical Risks Need Transparent Policies

On the ethics front, most respondents were optimistic about AI’s overall impact if newsrooms maintain standards. But 60% did express concerns about issues like algorithmic bias leading to inaccuracies or skewed coverage. Others pointed to risks surrounding misinformation and via “filter bubbles” from hyper-personalized content. The report advises transparency from newsrooms about their use of AI to maintain public trust. This could include flagging AI-written stories and explaining personalization algorithms. Ongoing training and dialogue about ethics is also recommended.

Imagining an AI-Powered News Future

Looking ahead, the respondents had ideas for AI applications in three timeframes. First, improving immediate efforts; second, innovating in 2-5 years with newer tools; and third, long-term rethinking of structures and processes. Asked what tools they most wanted, tagging/data extraction, automated content production, and personalized recommendation engines topped the list. But actually making this future happen will require major investments in staff training, recruitment, and cross-industry collaboration.

Overall, the Journalism AI Report provides a trenchant look at how news media is grappling with artificial intelligence. While AI is enhancing newswork already, the report urges responsibility. News organizations must optimize AI’s opportunities while ensuring quality journalism remains at the core. To make this a “revolution” not disruption means getting the AI fundamentals right in the next critical years.

This post was created with the assistance of Claude AI by Anthropic.

Share this Post: