AI Image Tools Accept Most Fake Political News Prompts
A test by AI startup Logically found that Midjourney, DALL-E 2, and Stable Diffusion accepted most prompts seeking fabricated political images.
Updated on Nov. 3, 2026 in Artificial Intelligence
Live Poll
Should AI image generators be required to block fake political images?
Logically tested over 40 political narratives across the U.S., U.K., and India and found that more than 85% of prompts to generate fake political news were accepted by the image generators. Midjourney produced the highest-quality images and had the most content moderation, while DALL-E 2 and Stable Diffusion had more limited safeguards.
Why it matters
The findings suggest that generative image tools could be used to produce misleading political content ahead of major elections in several countries. Logically urged social media platforms to adopt more proactive content moderation.
Logically tested over 40 political narratives and found that Midjourney, DALL-E 2, and Stable Diffusion together accepted over 85% of fake-news prompts. The test did not measure how often the resulting images were shared or how they were received by audiences.
The players
Logically
Logically is an AI startup that works with governments and companies to discover and manage online misinformation.
Midjourney
Midjourney is an AI image generator that produced the highest-quality images in the test and had the most content moderation policies in place.
DALL-E 2
DALL-E 2 is OpenAI's AI image generator, which had limited content moderation in the test.
Stable Diffusion
Stable Diffusion is an open image-generation model that accepted all prompts related to George Soros, Nancy Pelosi, and COVID announcements in the test.
The details
Logically tested prompts such as images of people stuffing ballot boxes tied to claims of a 'stolen election,' and a prompt replicating a viral Pentagon explosion image with Westminster Abbey as the target. DALL-E 2 and Stable Diffusion generated an image of Westminster Abbey in flames, while Midjourney rejected that prompt.
Timeline
2020: Trump lost the presidential election to Joe Biden, and 'stolen election' claims emerged.
May: An alleged Pentagon explosion briefly tanked the stock market.
Last month: DeSantis posted an AI-generated ad of Trump embracing Fauci.
Jan. 28, 2025: Deadline for the Conservative Party to call a UK vote.
2024: U.S. presidential election scheduled.
The Tech Race
The test compares the content moderation of Midjourney, DALL-E 2, and Stable Diffusion, showing that Midjourney had the strictest safeguards and produced the highest-quality images. The gap between these tools highlights a competitive difference in how image generators handle political misinformation.
Readers may encounter AI-generated political images that look authentic on social media, making it harder to verify what they see. Checking the source of striking political images before sharing them can help limit the spread of fabricated content.
The takeaway
Images that look realistic can spread quickly, so verifying a political image's origin before sharing it is a practical safeguard. Stronger moderation by both image tools and social platforms could reduce the risk ahead of major elections.
Further reading
Explore more on how these tools are shaping Artificial Intelligence coverage and safeguards.
Live Poll
Should AI image generators be required to block fake political images?
