This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A 1News investigation identified a network of automated Facebook pages, including 'NZ News Hub', that scrape legitimate news stories and use AI to rewrite them and generate unlabelled, inaccurate synthetic imagery. These pages impersonate real news outlets, dramatize emergency events, and have used manipulated photos of deceased individuals and grieving families. The activity has prompted warnings from civil defence and emergency management agencies due to the risk of undermining public trust in verified information during emergencies.
NZ News Hub, a Facebook page that reportedly imitated the style of a New Zealand news outlet, allegedly used AI to rewrite real news stories and pair them with unlabeled synthetic images. According to 1News, the page published manipulated depictions of real people, including a deceased minor and bereaved parents, and dramatized emergency-related scenes in ways that could distort public understanding of local events and undermine trust in legitimate news and emergency information.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Highly personalized AI-generated misinformation creating “filter bubbles” where individuals only see what matches their existing beliefs, undermining shared reality, weakening social cohesion and political processes.
Also classified as
The AI system generated highly inappropriate and offensive synthetic images, such as manipulating photos of a deceased minor and grieving parents without permission.
The automated generation of unlabelled, fabricated news stories and synthetic images at scale pollutes social media feeds and erodes public trust in legitimate journalism and official emergency communications.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The risk was caused by the intentional decision of human operators to set up fully automated pipelines that scrape news and generate unlabelled, misleading AI content for social media engagement.
Added categories peak at 2: Minor (Psychological Harm, Direct)
An automated network of AI-powered Facebook pages in New Zealand scraped and fabricated news stories, generating unlabelled synthetic images of disasters and deceased individuals. While causing localized distress and prompting warnings from emergency management agencies due to the risk of eroding public trust, the national security impact remains minor and manageable under standard procedures.
The report describes ChatGPT and Google's image generation tools being used as tools within a scripted, automated workflow. The systems perform specific functions (rewriting text and generating images based on direct inputs/prompts) without independent goal-directed planning or autonomous decision-making outside their programmed tasks.
Limited Risk: The report describes the generation of synthetic media and deepfakes without clear labelling. Under the EU AI Act, providers of AI systems that generate or manipulate image, audio or video content (deepfakes) must clearly disclose that the content has been artificially generated or manipulated.
The AI systems were integrated into an automated workflow to scrape news, rewrite it for social media, and generate accompanying synthetic images to populate Facebook pages.
The investigative report from 1News is highly detailed, providing specific examples of the pages, the number of posts reviewed (209), the specific AI tools used (ChatGPT, Google's tools via SynthID), and direct quotes from experts and emergency management agencies. The evidence of automation is clear, including raw prompts left in the posts.
None. The facts clearly point to an automated AI-driven misinformation/content farm operation.
Technology
Automated LLM Rewriting
Scripted workflows automatically rewrote real news stories using LLMs.
Synthetic Image Generation
AI tools generated realistic but fabricated images of emergency events.
Failure to Auto-Label AI
Platforms failed to automatically detect and flag unlabelled synthetic media.
Data Inputs
Scraped Legitimate News
Real police notifications and press releases were used as input data.
Unpermitted Personal Photos
Real photos of deceased individuals were used as inputs for AI manipulation.
Human Factors
Audience Verification Difficulty
Users struggled to distinguish fake pages from legitimate news rooms.
Intentional Deceptive Design
Operators designed pages to mimic trusted media brands to mislead users.
Process and Methods
Fully Automated Workflows
Content was scraped, rewritten, and posted without any human review.
Lack of Content Disclosure
No process was implemented to disclose the use of AI tools to users.
Management
Engagement-Driven Goals
Page operators prioritized social media engagement over factual accuracy.
No Editorial Oversight
The automated system lacked management-level editorial guidelines.
Regulatory Environment
Lack of AI Labeling Mandates
No regulations existed to mandate clear labeling of AI-generated content.
Weak Platform Accountability
Regulatory gaps allowed social media platforms to host fake news unchecked.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
