By using this site, you agree to the Privacy Policy and Terms of Use.
Accept
MediaBUZZ
MediaBUZZMediaBUZZ
Search
  • Latest Articles
    • Press Releases
    • Appointment & Announcements 2025
  • Latest News
    • Technology
    • Efficiency
    • Possibility
    • Ethicality
    • Creativity
  • Archive
    • EMags 2013
    • EMags 2014
    • EMags 2015
    • EMags 2016
    • EMags 2017
    • EMags 2018
    • EMags 2019
    • EMags 2020
    • EMags 2021
    • EMags 2022
    • EMags 2023
  • Advertising
  • Contact
Have an existing account? Sign In
Follow US
  • Home
  • Privacy Policy
  • Terms of Use
  • Imprint
  • Contact

© 2004-2023 Mediabuzz Pte Ltd. Reg. No. 200407301C | All Rights Reserved.
EthicalityTechnology

Manipulated language AI can write manifestos that appear human and radicalize automatically

MediaBUZZ
Published: October 24, 2025
SHARE

Shockingly, researchers at the Middlebury Institute of International Studies at Monterey, California, managed to teach Artificial Intelligence (AI) “GPT-3” to believe in the “QAnon” conspiracy theories that are widely spread on the Internet.

Generative Pre-trained Transformer 3 (GPT-3) is an autoregressive language model that uses deep learning to produce human-like text. It is the third-generation language prediction model in the GPT-n series created by OpenAI, a for-profit San Francisco-based AI research laboratory. It was introduced in May 2020 and has been in beta testing.

The quality of the text generated by GPT-3 is so high that it is difficult to distinguish from that written by a human, which has both benefits and risks, OpenAI researchers and engineers warned from the beginning.

“We experimented with the GPT-3 language model for several months to assess its possible abuse by extremists. We purposely built a Q-Bot for this,” says Alex Newhouse, co-author of the research report.

The scientists warn that manipulated language AI can write manifestos that appear human and thereby radicalize people.

In their experiment, the scientists fed GPT-3 with the views of the QAnon movement about a satanic “deep state”. They had previously asked the algorithm questions on this topic and received neutral, factual answers. After the learning process, the AI has formulated sentences on the same questions, the content of which could also come from human QAnon conspiracy theorists, proving that such algorithms can be easily radicalized.

According to Newhouse, machine learning makes it possible to program an extremely adjusted but emotionally appealing chatbot that spreads conspiracy theories on the Internet without much effort. Hence, he warns of the dangers of the next wave of synthetic text and appeals to lay down certain standards and regulations now.

Funnily enough, the majority of Republicans believe the far-right conspiracy theory “QAnon, according... responded that the QAnon conspiracy is “not true at all.”Tommy Beer and Jack Brewster, both writers at Forbes, state that “QAnon supporters claim President Trump is  from a cabal of Satan-worshiping pedophiles (consisting mostly of Hollywood celebrities, liberal politicians and “deep-state” government officials), who are running a secret child sex-trafficking ring, while “the FBI has identified QAnon conspiracy theorists as “extremists” who pose a potential .”Clearly, both algorithms and human beings can be fooled when fed with manipulatory information, but in this case, the machine is used to manipulate us.

MediaBUZZ LogoJoin MediaBUZZ

Subscribe to Continue Reading

Get exclusive AI insights for marketers and business leaders - newsletters, strategies, and expert tips included.

Subscribe for FreeAlready a Member? Login

Stay ahead in the AI-driven marketing world.

TAGGED:artificial intelligenceGPT-3Middlebury Institute of International Studies

Sign Up For Daily Newsletter

Be keep up! Get the latest breaking news delivered straight to your inbox.
By signing up, you agree to our Terms of Use and acknowledge the data practices in our Privacy Policy. You may unsubscribe at any time.
Share This Article
Facebook LinkedIn Email Copy Link Print
ByMediaBUZZ
Follow:
Independent online publisher in the Asia Pacific region focusing on the business of digital media and marketing.
Previous Article Man vs Machine: Will AI put our humanity at stake?
Next Article Marketers hardly use the potential of AI yet
Search
Popular Posts
  • AI Can Scale Content. But Can It Scale Meaning?

    September 11, 2026
  • Open or Closed AI? Marketing Has More at Stake Than It Thinks

    September 4, 2026
  • The AI Honeymoon Is Over. Where’s the ROI?

    September 11, 2026
  • When AI Does the Math, Who Checks the Numbers?

    September 4, 2026
  • When AI Acts for the Brand, Cybersecurity Becomes Marketing’s Problem

    September 1, 2026
Categories
Weather
29°C
Singapore
light rain
29° _ 28°
85%
3 km/h
Tue
33 °C
Wed
32 °C
Thu
29 °C
Fri
30 °C
Sat
30 °C

You Might Also Like

PossibilityTechnology

Consumer power climbs to new heights

5 Min Read
EfficiencyTechnology

Genesys invests in digitization and AI to provide intelligent and seamless service

4 Min Read
Archive 2023Technology

Alibaba Cloud unveils new AI Model to support enterprises’ intelligence transformation

46 Min Read
Archive 2023Technology

SentinelOne enhances cloud security with Snyk

5 Min Read
MediaBUZZMediaBUZZ
Follow US
© 2004-2026 Mediabuzz Pte Ltd. Reg. No. 200407301C | All Rights Reserved.
  • Home
  • Privacy Policy
  • Terms of Use
  • Imprint
  • Contact
Welcome Back!

Sign in to your account

Username or Email Address
Password

Lost your password?

Not a member? Sign Up