Financial Times warns AI written text is flooding daily communication

In short: The Financial Times says AI written text is spreading quickly across emails, marketing, and research, and society will need to adapt to both the benefits and the downsides.

What's going on

The Financial Times editorial board argues that computers are now producing large amounts of readable text on demand, from LinkedIn posts to company statements and even scientific papers. These tools are often called large language models, or LLMs (software trained on huge amounts of writing so it can predict the next word, like very advanced autocomplete).

After ChatGPT appeared in late 2022, researchers and readers tried to spot “tells” that suggest something was written by a machine. The editorial points to patterns such as repeated phrasing and certain fashionable words showing up more often. It cites a Barron’s analysis that found a sharp rise in a specific AI style of sentence structure in corporate documents during 2024.

The editorial also notes the growth of “AI humanisers”, tools that rewrite machine text to sound more human, sometimes by adding small imperfections. At the same time, some companies sell tools that claim to detect AI writing. The FT’s view is that chatbots can already imitate human writing convincingly in many cases.

What to watch

The editorial highlights three risks: more bland, copy and paste style writing as AI learns from AI, more machine written text being read mainly by other machines, and weaker human writing skills over time. It also argues there is no clear way back, since AI writing helps with spelling, grammar, and writing in a second language, and it can save time for professionals who need to publish a lot.

Source: Financial Times