Bankrate's Resumption of AI-Generated Content Allegedly Continuing to Produce Inaccurate and Misleading Information

Bankrate, and its sister site CNET, both owned by Red Ventures, resumed publishing AI-generated articles claiming thorough human fact-checking. However, new articles are alleged to contain numerous factual errors, including inaccurate statistics and misleading information. Despite public criticism, the company defended its use of AI and blamed out-of-date datasets for the errors. In addition to the errors, the incident raises questions about the ethical use of AI in journalism, especially given the company's insistence on "fact-checked" content.

Risk classification

• Primary risk domain: 3 Misinformation
• Primary risk subdomain: 3.1 False or misleading information

The AI system generated and published false economic and housing statistics, spreading incorrect information to readers seeking financial advice.

Additional risk subdomains

• 7.3 Lack of capability or robustness: The AI system failed to perform reliably by generating inaccurate data points, demonstrating a lack of capability in factual generation.
• 5.1 Overreliance and unsafe use: Human editors over-relied on the AI system and failed to perform the promised thorough fact-checking before publication.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The factual errors were generated by the AI system during its active deployment on Bankrate's live website, representing an unintended failure of the generation process.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: The system generates public-facing content, which falls under transparency obligations requiring users to be informed that the content is AI-generated.

AI system and alleged parties

• AI system: in-house natural language generation platform
• AI purpose: Writing Assistant; Automated Content Curation
• Behaviour type: Assistant
• Alleged developer: Red Ventures, Bankrate
• Alleged deployer: Red Ventures, Bankrate
• Alleged harmed parties: Journalistic integrity, General public

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. 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 Negligible
• 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 Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Epistemic

Reported: The report explicitly describes epistemic harm in the form of AI-generated articles containing factual errors regarding housing prices, salaries, and unemployment rates.

Directly caused: The AI system generated multiple incorrect data points about Boulder, Colorado, which were published on Bankrate's website.

Inferred additional harm: Unsuspecting readers searching Google for financial and relocation advice may have been misled by these false statistics, potentially affecting their financial planning or perceptions of these areas.

People affected

• Occurrences reported: 1

Potential causes

Management

• Prioritizing SEO over Accuracy: Management pushed AI content to capture search traffic without safeguards.
• Premature AI Resumption: Management restarted AI publishing before establishing robust reviews.

Technology

• Flawed Database Integration: AI tool pulled outdated data from internal database without validation.
• Lack of Automated Validation: No system to cross-reference generated text with live external sources.

Data Inputs

• Outdated Internal Dataset: The database used for the article contained outdated data from last year.
• Inaccurate Sourced Data: Data points for home prices and salaries did not match actual sources.

Human Factors

• Inadequate Editor Review: Editors failed to catch blatant factual errors in the AI-generated text.
• Overreliance on Automation: Staff assumed the AI-assist tool and database outputs were accurate.

Process and Methods

• Flawed Fact-Checking Process: Verification workflow did not require validation against primary sources.
• Hasty Publishing Workflow: Process prioritized rapid output over thorough content verification.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides highly specific, verifiable examples of the AI's factual errors, comparing them directly to the cited source databases. The publisher's response and explanation are also documented, leaving little ambiguity about the nature of the failure.
• Ambiguities identified: The exact mechanism of the data retrieval failure (whether the AI hallucinated or simply pulled from an outdated database as claimed) remains slightly disputed between the publisher and the reporter.
• Alternative interpretations: The publisher argues this was a database retrieval error rather than a generative AI failure, though the output was still presented as an AI-generated article.

The incident involved Bankrate publishing AI-generated financial articles containing minor factual and statistical errors due to outdated datasets and inadequate human oversight. This commercial publishing error poses negligible threats to national security, critical infrastructure, or sovereignty.

• Overall national security impact: Negligible
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents a resolved commercial event with no active or emerging national security crisis.
• Autonomy: Human-supervised. The AI system was designed to assist human writers and editors, though the promised human oversight failed to catch the errors before publication.
• Novelty: Established threat. AI hallucinations and factual errors in automated journalism are well-documented and have occurred in previous incidents.

Impact by dimension

• Physical security: Negligible. The incident involves local real estate and salary statistics on a consumer financial website, posing no threat to physical systems, critical infrastructure, or human safety.
• Information security: Negligible. While the incident involves misinformation, it was unintentional, commercial in nature, and lacked foreign adversary involvement or systematic intelligence compromise.
• Sovereignty: Negligible. The errors in consumer financial articles did not impact government decision-making, electoral systems, or state authority.
• Economic security: Negligible. Although the articles contained incorrect financial statistics, the impact was limited to minor consumer advice errors and did not threaten national financial systems or strategic industries.
• Societal stability: Negligible. The incident caused minor epistemic harm to readers of a financial blog but posed no threat to social cohesion, civil liberties, or population-scale stability.