Most content consumers spot misleading info from AI.

AI-written content is everywhere, and telling it apart from human writing is not always easy. A poll of 2,557 U.S. adults conducted by Pangram and YouGov revealed that 67% of online content consumers believe they have seen AI-generated content providing false, misleading, or incorrect information.

Younger web users notice this trend the most. About 73% of online Gen Zers say they have seen misleading AI material, with 55% seeing it frequently. Platform habits also play a role. Among users surveyed, 87% on Discord, 84% on Reddit, 78% on LinkedIn, 76% on X, 72% on Instagram and YouTube, 69% on Snapchat, 68% on TikTok, and 67% on Facebook reported spotting inaccurate AI posts.

This widespread bad content fuels deep distrust across the web. Overall, 69% of online users trust AI-generated material less than human-made pieces, while only 8% trust AI more. Skepticism peaks around sensitive topics:

  • News and current affairs: 68% trust human content more, compared to 8% for AI.
  • Opinion pieces: 69% trust human content more, versus 8% for AI.
  • Product reviews: 66% trust human content more, versus 9% for AI.
  • Legal documents: 66% trust human content more, versus 10% for AI.
  • Educational material: 59% trust human content more, while 12% trust AI.

Because of this gap, 61% of users say they avoid engaging with content if they think it comes from AI. Yet spotting it remains a major hurdle. About 35% of adults do not believe they can definitively tell AI writing apart from human writing. Gen Z feels most confident, with only 22% doubting their detection skills, compared to 46% of baby boomers.

Despite these detection challenges, 71% of web users believe they encounter AI content at least a few times a week, with 41% seeing it daily. Surprisingly, 48% of baby boomers report seeing AI content daily, higher than Gen Z at 37%.

With undisclosed AI cluttering feeds, 75% of online users say they want a tool or browser extension that flags AI material. Detection options like Pangram aim to clear up the confusion by identifying generated text with high accuracy, helping people navigate the web with confidence.

Source: Pangram

Leave a Reply

Your email address will not be published. Required fields are marked *