In Brief: Falsehoods Seen as a Global Threat; Legal Releases Cliché-Ridden

In 25 countries around the world, adults view the spread of false information online as a major threat, Pew Research Center has found. People also see terrorism and the condition of the global economy as major threats, according to Pew’s analysis.

A median of 72% of adults across 25 countries called the spread of false information online a major threat to their country. The median is the middle number in a list of all percentages, from highest to lowest. Some 21% of respondents consider the spread of online falsehoods a minor threat, while 5% say it isn’t a threat.

In seven countries — the United States, the United Kingdom, Germany, the Netherlands, Poland, Sweden, and South Korea — people rank the spread of false information online as more of a threat than any other issue they were asked about in the survey.

Across all countries surveyed, a median of 70% of adult respondents see the condition of the global economy as a major threat. Another 27% say it’s a minor threat, while 4% say it’s not a threat.

Legal Press Releases Lean on the Same Clichés, Analysis Shows

An analysis of press releases issued by law firms over the last year finds what it calls “the most tired, overused words and phrases” in those releases. 

Infinite, a communications firm, analyzed 479 press releases issued by U.S. and U.K. corporate law firms from June 2024 to July 2025. Some 87% included at least one of what Infinite calls the 10 most commonly overused words: “leading,” “best,” “most,” “pleased,” “excited,” “delighted,” “proud,” “excellence,” “thrilled,” and a tie between “largest” and “innovative.”

When using such words in press releases or other communication, “Moderation is key,” said Matthew Gilleard, Infinite’s associate director. “Just as overuse erodes impact, sparing use increases it,” he said. 

Audiences such as journalists and corporate counsel may respond with skepticism when reading announcements that contain many superlatives and overused words. Said Infinite director Tal Donahue, content grounded in facts and optimized for internet and AI discovery “is essential for any business that wants to boost visibility, trust and authority online.” 

Why People Collect, and What It Means for Brands

Collecting has long been a universal human behavior. People might collect because of a psychological need to control their environment, research from the University of Arizona and Lingnan University in Hong Kong suggests. And turbulent times such as the COVID-19 pandemic tend to increase the appeal of collecting, the researchers found.

Asked to imagine collecting vinyl records, study participants who had a higher desire for control were willing to spend more money or time to complete their collections. 

The research has implications for brands. Companies might succeed by marketing products that are part of completable sets, the study suggests. The popular Pokémon franchise, which has a reported market value of more than $100 billion, features the tagline “Gotta Catch ’Em All.”

“Structure is what makes collecting so powerful,” said Martin Reimann, an associate professor of marketing at the University of Arizona. But he cautioned that by always adding new collectibles, brands might frustrate consumers who thought their sets were complete.

Study: Chatbots Sway Political Views

All AI models are biased, and biased AI chatbots can influence people’s political views with just a few messages, a recent study finds. 

Researchers at the University of Washington and Stanford University recruited self-identifying Democrats and Republicans to form opinions on obscure political topics. Participants were randomly assigned three versions of the AI chatbot ChatGPT: a neutral model, a version with liberal bias, and a version with conservative bias.

In the study, both Democrats and Republicans were more likely to politically lean in the direction of the biased chatbot after interacting with it. Participants from both parties leaned further to the political left after talking with a liberal-biased system. In another recent study, participants thought that ChatGPT and all large language AI models lean liberal.

Explicitly biased chatbots often tried to persuade users by altering how they framed topics. Participants with higher self-reported knowledge of AI shifted their views less significantly after talking with biased chatbots, the researchers found.