Chatbots Spread More False News in 2025, Researchers Find 😱

Welcome to another edition of Horizon AI,

It’s long been reported that you can’t trust chatbots to deliver accurate information on current news, but a recent study shows the misinformation problem is not improving. In fact, it has gotten much worse.

Let’s jump right in!

Read Time: 4.5’ min

Here's what's new today in the Horizon AI

• Chart of the week: 17 ChatGPT Prompt Engineering Techniques

• Study Finds Chatbots Are Now Twice as Likely to Share Misinformation

• Free Resources

• AI tools to check out

• Video of the week

TOGETHER WITH ROKU

Find your customers on Roku this Black Friday

As with any digital ad campaign, the important thing is to reach streaming audiences who will convert. To that end, Roku’s self-service Ads Manager stands ready with powerful segmentation and targeting options. After all, you know your customers, and we know our streaming audience.

Worried it’s too late to spin up new Black Friday creative? With Roku Ads Manager, you can easily import and augment existing creative assets from your social channels. We also have AI-assisted upscaling, so every ad is primed for CTV.

Once you’ve done this, then you can easily set up A/B tests to flight different creative variants and Black Friday offers. If you’re a Shopify brand, you can even run shoppable ads directly on-screen so viewers can purchase with just a click of their Roku remote.

Bonus: we’re gifting you $5K in ad credits when you spend your first $5K on Roku Ads Manager. Just sign up and use code GET5K. Terms apply.

Use code GET5K now

Chart of the week

17 ChatGPT Prompt Engineering Techniques

AI News

AI & MEDIA

Study Finds Chatbots Are Now Twice as Likely to Share Misinformation

AI assistants are now twice as likely to spread false information as they were a year ago, according to a recent NewsGuard study.

Details:

• Researchers found that the top ten chatbots repeated misinformation about current news topics in 35% of tests, up from 18% in 2024.

• Model results varied widely, with Inflection performing worst at 56.67% and Perplexity at 46.67%, while Claude and Gemini did best at 10% and 16.67%, showing that design choices and guardrails still matter a lot.

• The rise is linked to chatbots adding real-time web search and relying on a web saturated with low-quality and unreliable sources.

• Another consequence of these web search capabilities is that chatbots have stopped refusing to answer questions, with the denial rate dropping from 31% in August 2024 to zero a year later, which worsens the misinformation problem.

• Perplexity’s swing is striking, falling from a perfect 100% debunk rate in August 2024 to echoing false claims nearly half the time only one year later.

Chatbots are useful for speed and summaries, but they remain far from reliable when it comes to delivering the truth, especially about recent news. Verifying claims against trusted outlets is still essential.

Source

AI Findings/Resources

🍌 Turn your photos into images of custom miniature figures in just one prompt with Nano Banana.

🎵 The catchiest songs of all time, according to AI

📈 The Economist partnered with a research and AI firm to measure corporate culture across 900 companies in 19 industries

⭐ An AI-generated 1960s-style version of 50 Cent’s ‘Many Men’ is going viral

⚠️ Safety of AI chatbots for children and teens faces US inquiry

AI Tools to check out

🎬 Boba Anime 1.3: An AI model built for anime storytelling, from stills to cinematic reels.

⭐ Drawnaq: AI-powered design platform.

⚡ Framer: Turn ideas into high-performing websites, fast.

🦾 Masonry: Collaborate with multiple AI agents to automate workflows, assign tasks, and boost business productivity with seamless AI-powered automation.

👉 Scrybe: Grow your LinkedIn following in less than 5 minutes a day.

Video of the week

What Jobs Will Remain in 2030 After AGI?

Dr. Roman Yampolskiy, an expert in AI safety, discusses his concerns about the rapid and unregulated advancement of AI.

He argues that Artificial General Intelligence is only a few years away and will have a significant impact on our everyday lives, with the job market hit the hardest. He warns this could lead to an unemployment rate of up to 99% as AI takes over most jobs.

The conversation also touches on other interesting topics, including the simulation hypothesis, with Dr. Yampolskiy stating he is "very close to certainty" that we are living in a simulation.

Some timestamps (you can find more in the video description):

02:28 How to stop AI from killing everyone

04:35 What's the probability something goes wrong?

04:57 How long have you been working on AI safety?

11:38 What jobs will actually exist?

14:27 Can AI really take all jobs?

18:49 What happens when all jobs are taken?

22:04 Prediction for 2030

23:58 What happens by 2045?

28:51 Is anything more important than AI safety right now?

30:07 Can't we just unplug it?

37:20 What is most likely to cause human extinction?

39:45 No one knows what's going on inside AI

42:32 Thoughts on OpenAI and Sam Altman

56:10 Are we living in a simulation?

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻