September baseline across more than 2,100 Medicare Plan-ID entities finds major differences between Google AI and Microsoft Copilot in plan-fact coverage and exposed source attribution.
-- PRESCOTT ARIZ., Sept. 28, 2026 — A new analysis of more than 2,100 Medicare Plan-ID entities finds that observed source attribution in AI-generated plan information is concentrated among a relatively small group of private publishers—and that Google AI and Microsoft Copilot differ substantially in both the Medicare facts they assert and the extent to which they expose source attribution.
The Medicare Visibility Report: September 2026 Baseline, published by Trust Publishing Institute and MedicarePlans.com, examined 39 standardized categories of Medicare plan information across approximately 2,100 CMS Plan-ID entities. Rather than treating rankings or citation totals as a complete measure of AI visibility, the study measured whether an observed answer engine asserted a specific plan fact and, when exposed, which source it associated with that assertion.
Across the measured population, the three largest sources accounted for approximately 70% of observed Google AI attribution and 89% of observed Microsoft Copilot attribution. The top 10 sources accounted for approximately 91% of Google AI attribution and 99% of Copilot attribution.
Google AI exposed attribution from 188 source domains, while Copilot exposed attribution from 74. The findings show that more observable sources do not necessarily mean attribution is broadly distributed.
“Search engines discover documents; answer engines assemble assertions,” said David W. Bynon, researcher and author of the report. “A ranking or citation count cannot, by itself, show which facts an answer engine presents about a Medicare plan or which publisher is associated with the evidence for each fact. This baseline is designed to measure those relationships directly.”
The findings come as TheStreet independently reported that AI systems can produce materially different Medicare answers based on seemingly minor differences in consumer language and examined research from the Medicare Visibility Monitor. Related reporting: Medicare shoppers face a hidden AI problem this open enrollment
Google and Copilot Show Different Profiles
The report found that, under the study’s standardized Plan-ID-specific protocol, Google AI asserted approximately 54% of the 39 predefined plan facts, compared with 68% for Microsoft Copilot.
However, Google AI provided explicit source attribution for about 94% of its asserted facts, compared with about 50% for Copilot.
The report calls these distinct measures knowledge completeness and knowledge provenance:
Knowledge completeness is the share of the 39-fact model explicitly asserted in an observed response.
Knowledge provenance is the share of asserted facts accompanied by identifiable source attribution.
Neither measure establishes factual accuracy, consumer exposure, source originality, or the reason an answer engine selected a source.
The systems also differed sharply by fact category. Copilot asserted emergency-room copay information for 90.8% of measured Plan-ID entities, compared with 24.8% for Google AI. Copilot asserted Medicare eligibility for 96.9% of plans and residency requirements for 89.8%, compared with 38.2% and 36.9%, respectively, for Google AI. Google AI more frequently asserted service-area information and insulin-cost information.
Attribution Varies by Engine and Fact
Medicare.org was the largest measured source of attributed assertions on both observed systems. On Google AI, Medicare.org was associated with 34,607 attributed assertions, followed by Q1Medicare with 20,911 and MedicareAdvantage.com with 14,225.
On Microsoft Copilot, Medicare.org was associated with 22,816 attributed assertions, followed by MedicareAdvantage.com with 7,710 and MedicarePlans.com with 3,234.
The report emphasizes that observed attribution is not a quality score and should not be interpreted as proof that a fact is accurate, current, original to a cited publisher, neutral, or preferred by consumers. One assertion may have multiple exposed source relationships, and an engine may assert a fact without exposing a source.
The report’s core measurement relationship is:
Plan-ID→assertion→engine→attributed source
This framework distinguishes among conventional search discovery, assertion production, exposed source attribution, and publisher-level evidence participation.
Medicare Information Infrastructure
CMS is a primary official source of Medicare plan data and plan-comparison information, while Medicare carriers provide plan-specific materials and operating information. The report finds that answer engines often attribute information to private publishers that organize, interpret, or republish Medicare information.
The findings raise an infrastructure question: how can authoritative Medicare information be structured and distributed so AI systems can reliably identify the correct plan, plan year, geography, eligibility requirements, costs, and benefits without unnecessary dependence on intermediary interpretations?
The report does not attempt to establish why an AI system selected a given source or whether a particular publishing method caused an attribution outcome. It measures observed answer-engine behavior under a defined protocol.
Consumers should treat AI-generated Medicare answers as a starting point for research, not as the final authority for an enrollment decision. Before enrollment, consumers should verify plan year, service area, eligibility, costs, drug coverage, provider participation, and key benefits through official Medicare resources and materials supplied directly by the plan.
The full report and study methodology are available at https://www.medicareplans.com/research/medicare-ai-visibility-study/ as a PDF download.
About The Report
The Medicare Visibility Report is an independent measurement project of Trust Publishing Institute and MedicarePlans.com. Research, methodology, analysis, and reporting were conducted by David W. Bynon.
Bynon manages publishing technology and editorial systems for Medicare.org, a publisher included in the measurement. Health Network Group LLC, an Allstate company, owns and operates Medicare.org. Neither Medicare.org nor Allstate commissioned the report or determined its methodology, findings, or conclusions. Observed publishers were evaluated using the same measurement and aggregation rules, and assertion-level provenance was retained for auditability.
The September 2026 baseline does not establish causality, factual accuracy, consumer exposure, referral traffic, commercial influence, or cross-engine semantic agreement.
Contact Info:
Name: David W Bynon
Email: Send Email
Organization: Trust Publishing Institute
Address: 1800 Club House Drive #93, Bullhead City, AZ 86442, United States
Website: https://trustpublishing.org/
Source: PressCable
Release ID: 89204599
Should you identify any discrepancies, concerns, or inaccuracies in the content provided in this press release or require assistance with a press release takedown, we strongly urge you to notify us promptly by contacting [email protected] (it is important to note that this email is the authorized channel for such matters, sending multiple emails to multiple addresses does not necessarily help expedite your request). Our responsive team is committed to addressing your concerns within 8 hours by taking necessary actions to resolve identified issues diligently or guiding you through the necessary steps for removal. Our dedication lies in providing accurate and reliable information.
