Hoodline Accused of Misleadingly Attributing AI-Generated Articles to Human Authors

In 2023, the news site Hoodline is reported to have begun publishing AI-generated articles with fake bylines, headshots, and biographies, allegedly misleading readers into believing they were authored by real journalists. This practice diminishes public trust and exemplifies the potential dangers of AI in journalism. Despite a disclaimer, the use of AI was not transparent.

Risk classification

• Primary risk domain: 6 Socioeconomic & Environmental
• Primary risk subdomain: 6.3 Economic and cultural devaluation of human effort

The incident highlights the displacement of human journalists and the devaluation of original reporting through the automated rewriting of existing content by AI.

Additional risk subdomains

• 4.3 Fraud, scams, and targeted manipulation: The use of fake human profiles, biographies, and AI-generated headshots was designed to deceive readers into believing the content was written by real local journalists.
• 3.2 Pollution of information ecosystem and loss of consensus reality: The systemic generation of low-quality, automated articles across a national network of local sites threatens to pollute the local news ecosystem.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The deception and copyright issues were caused by the intentional decisions of Hoodline's management to deploy AI-generated content under fake human personas post-deployment.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: Limited Risk. The system generates text content and previously used synthetic headshots, which triggers transparency obligations under the EU AI Act to ensure users are informed that they are interacting with AI-generated content.

AI system and alleged parties

• AI system: proprietary and custom-built software
• AI purpose: Writing Assistant; Automated Content Curation
• Behaviour type: Assistant
• Alleged developer: Hoodline
• Alleged deployer: Hoodline
• Alleged harmed parties: Journalism, Hoodline readers, General public

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Severe. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Minor
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Severe, indirect Substantial
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: The report does not explicitly quantify specific financial losses for the occurrences, though it notes copyright concerns and job displacement.

Indirectly caused: Human journalists lost employment and income as Hoodline transitioned from a human-staffed newsroom to AI-generated content.

Inferred additional harm: The displacement of dozens of local journalists likely resulted in significant aggregate loss of income, estimated in the hundreds of thousands of dollars.

Epistemic

Reported: Yes, the report explicitly describes concerns that relying on AI could supercharge the spread of misinformation and notes that other outlets using similar AI made embarrassing factual errors.

Directly caused: The deployment of AI-generated articles under fake human personas misled millions of readers into believing they were reading original reporting by local human journalists.

Indirectly caused: The practice threatens to diminish overall public trust in journalism and news organizations.

Inferred additional harm: The rewriting and aggregation of content by AI without original reporting likely spreads unverified or slightly distorted information across multiple local sites.

People affected

• Occurrences reported: 1
• People reportedly harmed: 50
• People reportedly exposed: 2000000

Potential causes

Management

• Prioritizing Cost and Traffic: Focusing on scaling traffic twentyfold and reducing costs over quality news.
• Replacement of Human Newsrooms: Replacing local human journalists with AI personas to cover news deserts.

Technology

• Proprietary AI Article Generation: Custom software combined with AI partners to write publish-ready articles.
• AI-Generated Human Personas: Creation of fake headshots and biographies to mimic real human journalists.

Data Inputs

• Unpermitted Content Harvesting: Rewriting and aggregating press releases and articles from other outlets.

Human Factors

• Deceived Public Perception: Readers were misled into believing articles were written by local journalists.
• Decreased Public Trust: The public is less trusting of news labeled as AI or written by fake personas.

Process and Methods

• Superficial Transparency Disclosures: Using a small AI badge next to fake human names instead of full disclosure.
• Removal of Archived Evidence: Requesting removal of Wayback Machine archives to hide past practices.

Regulatory Environment

• Potential Copyright Infringement: Using other news organizations' content without permission or compensation.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides clear, detailed accounts from multiple sources (CNN, former journalists, media experts, and the company's CEO) regarding the deployment of AI, the use of fake personas, and the resulting public backlash and ethical concerns. The facts of the deployment are uncontested, though the company defends its practices.
• Ambiguities identified: The specific proprietary AI models and partners used by Hoodline are not disclosed.
• Alternative interpretations: The company frames the AI as a tool to save local journalism in news deserts, whereas critics view it as a deceptive cost-cutting measure that devalues journalism.

A local US news network transitioned to AI-generated articles using fake human personas and synthetic headshots. While raising significant ethical concerns regarding media trust, transparency, and journalist displacement, the incident carries negligible direct national security implications.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Impress3 Media

Threat characteristics

• Imminence: Long-term. The situation represents an ongoing strategic and ethical concern regarding media integrity rather than an active, imminent national security crisis.
• Autonomy: Human-supervised. The AI systems operated under human supervision, generating draft articles that were published under the management direction of the parent company.
• Novelty: Evolved capability. Represents an evolution of existing automated content generation techniques, combining text generation with synthetic personas to mimic local newsrooms.

Impact by dimension

• Physical security: Negligible. No physical systems, infrastructure, or human safety concerns were impacted by the deployment of AI-generated local news articles.
• Information security: Minor. The incident involved systemic generation of articles under fake personas, contributing to information pollution, but lacked state-actor involvement or targeted national security disinformation.
• Sovereignty: Negligible. No core government functions, sovereignty decisions, or constitutional processes were targeted or disrupted by this commercial news automation.
• Economic security: Negligible. While local journalists experienced job displacement, the incident does not pose a threat to strategic industries, critical supply chains, or national economic stability.
• Societal stability: Minor. The deployment of fake human personas and AI-generated content devalues human effort and risks eroding public trust in local journalism, presenting minor societal concerns.