Strategic PartnersIssue 083 records
A day after the Nov. 4 walk-back coverage, independent reporting confirms Google has not softened the K-12 policy further: schools may keep Search without Gemini apps, but Search still embeds AI (Overviews/AI Mode or Guided Learning)—there is no AI-free school Search (Ars Technica, Nov. 5). Separately, Google began an extended global rollout of Guided Learning attached to Classroom assignments (Nov. 4 Workspace Updates), which requires Gemini On for the relevant OUs—tightening the link between chatbot enablement and the “responsible learning” path. Google’s Public Sector Summit recap (datePublished Nov. 5) is vendor marketing: it names Connecticut and Case Western among collaborators and highlights Wiz for Gov AI agents plus Gemini enterprise search on air-gapped Distributed Cloud, without measured SLED outcomes. 3 sources; no new HHS, public-safety, or local-government deployment evaluations.
- Evidence records
- 3
- Cross-source patterns
- 1
- Evidence classes
- 2 vendor claim1 independent reporting
- Outcomes
- 2 emerging1 cautionary
- Source freshness
- 3 new this fortnight
- Research completed
- 2026-11-06
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
Guided Learning Classroom path needs Gemini On; Search-on still keeps AI
Ars Technica confirms schools cannot get AI-free Search, while Workspace Updates makes Classroom Guided Learning assignments depend on Gemini being On for the OU. Interpretation: districts seeking a “responsible learning” AI path must enable Gemini for Classroom features, and districts that only enable Search still inherit AI Overviews/AI Mode.
Operating questionHow should a district’s AI policy treat Gemini enablement required for Guided Learning versus Search-on settings that still include AI?
Supporting evidenceArs TechnicaGoogle
Evidence ledger
Full record · every source keeps its link and limitations
Google confirms K-12 policy unchanged: schools can skip Gemini apps but not AI inside Search
Ars Technica reports a Google spokesperson said the company has not changed the new K-12 AI policy announced this week. Schools can disable Gemini apps while leaving Search active, or turn both off; enabling Search still means accepting AI Overviews/AI Mode or Guided Learning. Google will not offer AI-free school Search. The piece notes DOJ search-antitrust remedies may eventually strengthen alternatives.
Why it matters, evidence and limitations
- Why it matters
- K-12 districts and state education agencies using Google Workspace for Education must translate board AI bans into concrete admin settings knowing Search cannot be AI-free. Procurement and policy teams should document that “no Gemini” is not the same as “no AI in Search.”
- Evidence and measured results
- Primary evidence is Google spokesperson confirmation to Ars plus interpretation of Google’s blog and social clarification. No quantitative district outcomes. Corroborates the archived Search Engine Watch/NBC line that AI remains inside school Search.
- Limitations and uncertainty
- No admin-console verification in the article; behavior is as announced. DOJ index-sharing benefit to rival search is prospective, not current.
Google rolls out Guided Learning on Classroom assignments globally, requiring Gemini On
Google Workspace Updates states educators can add Guided Learning to Classroom assignments in a global extended rollout that started November 3, 2026 for Education Fundamentals, Standard, and Plus. Sessions can use Classroom sources as context; student UI hides Gemini navigation and disables new chats in-tab; teachers get class/individual insight summaries and transcripts. Admins must have Gemini set to On for the teacher/student group or OU.
Why it matters, evidence and limitations
- Why it matters
- U.S. K-12 districts evaluating Guided Learning as the “safer” AI path need to know Classroom assignment integration is shipping now and is gated on Gemini enablement—the same control under political scrutiny.
- Evidence and measured results
- Vendor product bulletin with rollout pace and availability tiers. Claims grounding in LearnLM research but this page does not present independent measured learning gains.
- Limitations and uncertainty
- Vendor claim; learning-science benefits are asserted, not independently validated here. Feature invisible until Gemini is On.
Google Public Sector Summit recap highlights Gemini for government, Wiz for Gov agents, and air-gapped GDC search
Google’s Nov. 5 Public Sector Summit recap says 2,700+ government and partner attendees heard updates including Gemini for government services, Wiz Red/Green/Blue agents now in Wiz for Gov’s FedRAMP High environment, and enterprise search capabilities for Gemini on Google Distributed Cloud for air-gapped deployments. Named collaborators include the State of Connecticut and Case Western Reserve University among others; no outcome metrics are provided.
Why it matters, evidence and limitations
- Why it matters
- Signals roadmap items state/local and higher-ed buyers may hear in partner conversations—governed agents, FedRAMP High security agents, and air-gapped Gemini search—without proving SLED deployments.
- Evidence and measured results
- First-party marketing blog with named customers and product announcements. No independent confirmation or measured benefits in this source.
- Limitations and uncertainty
- Vendor marketing; Connecticut and Case Western are named without describing scope, contract, or results. Federal/defense speakers dominate the narrative.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Vendor claim
- A supplier-provided assertion that has not been upgraded to independent evidence.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.