Skip to content
GovSmith

Privacy Policy

Effective October 31, 2026 · Replaces the version of October 21, 2026 · Terms of Service

This Privacy Policy explains how Blacksmith AI LLC DBA "GovSmith" ("GovSmith", "we", "us" or "our") collects, uses, shares and protects information when you use GovSmith: the client portal at app.govsmith.com, the work our staff do for your company, our website at govsmith.com and our related services (the "Service"), and the choices and rights you have.

The short version: we collect what we need to serve your company, our staff use AI tools on your content only to do the work you engaged us for, we do not sell or share personal information for advertising, we do not use your data to train AI models, and you can download your data and ask us to delete it.

1. Scope and our roles

This Policy covers the Service, our website at govsmith.com, our public legal pages, and the emails we send. It does not cover how we handle the personal data of our own employees and contractors as staff, which their agreements with us cover.

We act in two roles:

  • As a controller for the personal data we need to run our business: account and contact details, billing, security logs, support requests, and messages sent through our website.
  • As a processor (a service provider) on behalf of each client company for the Client Content its people share with us or that we prepare for it: documents, company profile data, pursuits, Briefs, Deliverables, messages, contacts and connector data. The client company decides why that data is collected; we process it under our Terms of Service, its Service Agreement and this Policy.

If you are on a client company's team, its Owner and Admins manage its people in the Service, and the vCMs assigned to it work with the data it shares. Questions about how your company uses the Service should go to your Owner.

2. Information we collect

Information you give us

  • Account: first and last name, email, username, password (hashed, never stored in clear), profile photo, job title, and the multi-factor or passkey credentials you register (we store a public key or an encrypted authenticator secret, never a private key).
  • Company profile: company name, website and logo, addresses, UEI and CAGE codes, NAICS and PSC codes, certifications, past performance, key documents, partners and contacts.
  • Client Content: files you upload or import, files you share with your vCM, messages to your vCM, meeting requests, approvals, pursuits, the submissions you record, and anything else you enter.
  • Connector data: when you connect Google Drive, OneDrive, SharePoint or another app, the files and data those connectors are permitted to read within the scopes you approve, and the files you choose to import.
  • Team and billing: the names and emails of people you invite, who may approve quotes, your acceptance of the Service Agreement and of each Quote, and your billing address. Card and bank details are entered with Stripe and never reach our servers.
  • Support and website: bug reports and support requests (with the page you were on and your browser details), and the discovery inquiries and demo requests you send through the forms on govsmith.com.

Information we create while doing the work

  • Briefs, recommendations and win estimates about opportunities for your company; Quotes; Deliverables; and the notes, internal estimates and time records our staff keep about the engagement, some of which are not shown in the Portal.
  • The GovSmith AI conversations, workspaces and memories our staff use to do your work.

Information collected automatically

  • Usage: the pages and features you use, the delivery time used, and timestamps. This powers Settings, Usage and Billing.
  • Device and connection: IP address, browser and operating system, screen size, language, and approximate location derived from IP.
  • Security logs: sign-ins and sign-outs, failed attempts, MFA and passkey events, session records, and administrative actions, kept in an audit log.
  • Error reports: when something breaks, an error report with your user ID and organization ID, and for some sessions a replay in which all text, inputs and media are masked.
  • Cookies and browser storage: see section 7.

Information from others

  • Sign-in providers: if you sign in with Google or Microsoft we receive your name, email and profile photo.
  • Payment processor: Stripe tells us whether a payment succeeded and the brand and last four digits of the card or bank account.
  • Public sources: company details from SAM.gov and USAspending, opportunity data from the government sources in section 5, and company logos looked up from a website domain.
  • Your company: the Owner or an Admin gives us your name, email and role when they invite you.

3. How we use information and our legal bases

We use information for the purposes below. Where the GDPR or UK GDPR applies, the legal basis for each is shown.

PurposeExamplesLegal basis
Provide the ServiceRun the Portal; assess opportunities and prepare Briefs, Quotes and Deliverables; organize and search your files; sync connectors; schedule meetingsPerformance of a contract
Accounts and securityAuthenticate you, enforce MFA and roles, limit staff access to assigned clients, detect abuse, keep audit logsContract; legitimate interest in security
BillingCharge onboarding, monthly and order fees; issue invoices; track included delivery timeContract; legal obligation
SupportAnswer requests and bug reports, reproduce problemsContract; legitimate interest
ReliabilityFind and fix errors using error reports and masked session replays (Sentry)Legitimate interest
CommunicationsSend email about your account, team, engagement, billing and security; announcements you can turn offContract; consent for marketing
ComplianceRespond to lawful requests, enforce our Terms, meet export control, sanctions and tax rulesLegal obligation; legitimate interest

We do not use Client Content to build profiles for advertising, and we do not make automated decisions about you that have legal or similarly significant effects.

4. How AI processing works

  • The Portal has no AI chat. Our staff use AI tools, including GovSmith AI, on your Client Content to do the work you engaged us for. When they do, the request, the relevant context (documents, company profile, pursuit details) and their instructions are sent to an AI provider to produce a result. The providers and their purposes are in the table in section 5.
  • Model calls go through OpenRouter, which forwards each one to the provider that runs the model; if OpenRouter is unavailable, we may call Anthropic or OpenAI directly on our own accounts. OpenRouter does not store prompts or answers, and we send it a one-way hash of the organization ID, never a name or an email.
  • Work on solicitations and your company's documents, including document extraction and the search index over your files, runs only on Anthropic and OpenAI. Anthropic and OpenAI keep API data for a short period to detect abuse, under their own terms, and do not train on it. Every other model runs only through providers that keep no data after answering (zero data retention).
  • We do not use your personal data, Client Content or AI outputs to train or fine-tune large language models, ours or anyone else's, and our settings with OpenRouter forbid training on them. This statement is made for the purposes of state privacy laws that require it, including Connecticut's.
  • Files are split into passages and turned into embeddings (numeric representations) so the Service can find relevant passages. Embeddings are stored in our database, each marked with the client company it belongs to and kept apart by row-level security.
  • GovSmith AI can keep short memories for the staff member using it, so later work can use them. Staff can see and delete them in Settings, Data controls, and turn the feature off.
  • When GovSmith AI searches the web, the search query goes to the model provider's own search tool or to OpenRouter's web search tool (powered by Exa).
  • Read-aloud sends only the text of a reply to ElevenLabs through OpenRouter; dictation runs in the browser.
  • Our staff review AI output before it reaches you in a Brief or a Deliverable, but it can still be wrong. The Terms of Service explain your responsibilities.

5. Who we share data with

We do not sell personal information, and we do not share it for cross-context behavioral advertising. We have not done so in the preceding 12 months.

We share information only as follows:

Subprocessors and service providers

The vendors below process data for us under terms that limit their use of it to providing their service, require confidentiality and security, and do not allow them to train AI models on it. We will update this table at least 30 days before a new subprocessor begins processing Client Content; client companies may object in writing on reasonable grounds. OpenRouter's own subprocessors are listed at trust.openrouter.ai.

VendorPurposeDataRegion
OpenRouterRoutes AI model requests to the provider that runs the model, including its web search tool (powered by Exa) for models without their own searchPrompts, documents and answers in transit, not stored; request metadata (model, provider, tokens, cost, timing) and a one-way hash of the organization ID, never a name or an emailUnited States (Google Cloud US regions)
AnthropicAI models (Claude) for GovSmith AI, proposal writing, document extraction and review, called through OpenRouter on our own Anthropic account (directly, if OpenRouter is unavailable)Prompts, documents and context our staff submit for AI processing; Anthropic keeps API data for a short abuse-monitoring period under its terms and does not train on itUnited States
OpenAIAI models (GPT) and the embeddings that make your files searchable, called through OpenRouter on our own OpenAI account (directly, if OpenRouter is unavailable)File text and search queries (embeddings); prompts and documents our staff submit; OpenAI keeps API data for a short abuse-monitoring period under its terms and does not train on itUnited States
Google (Gemini)AI models (Gemini) when our staff pick one in GovSmith AIPrompts, documents and context our staff submit for AI processing; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
xAIAI models (Grok) when our staff pick one in GovSmith AIPrompts, documents and context our staff submit for AI processing; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
PerplexityWeb-grounded research (Sonar) in GovSmith AIResearch questions and the context needed to answer them; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
Mistral AIAI models (Mistral) when our staff pick one in GovSmith AIPrompts, documents and context our staff submit for AI processing; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
CohereAI models (Command) when our staff pick one in GovSmith AIPrompts, documents and context our staff submit for AI processing; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
Open-model hosts on OpenRouter (such as Together AI, DeepInfra and Fireworks AI)Open-weight models (Llama, DeepSeek, GLM, Kimi, Nemotron, Phi) when our staff pick one in GovSmith AI; OpenRouter sends each request only to a host that keeps no dataPrompts, documents and context our staff submit for AI processing; through OpenRouter with zero data retention, so nothing is kept after the answerWhere the provider runs the model (no data retained)
ElevenLabsText to speech when a staff member plays a GovSmith AI reply aloud, through OpenRouterThe text of the reply being read aloud; zero data retentionWhere the provider runs the model (no data retained)
Amazon Web ServicesHosting and delivery of the app and the API (CloudFront, Lambda, S3, SQS and worker containers started on demand), email (Amazon SES), encryption keys (KMS), secrets, logs, malware scanning of uploaded files and threat detection (GuardDuty), encrypted backups, place lookups (Amazon Location Service), and the website govsmith.com with its form records and bot protection (AWS WAF)All Service data at rest and in transit, including uploaded files, the email we send, encrypted database backups, AI request metadata (never prompts or answers) and the addresses we look upUnited States (us-east-1)
NeonPrimary database (PostgreSQL with pgvector), including the search index over your files, with point-in-time restoreAccount, company, engagement, quote, order, pursuit, contact, message and audit records; file text, embeddings and short passages used for searchUnited States (AWS us-east-1)
StripePayments and invoices for onboarding, the monthly relationship and ordersName, email, billing address and payment method (card and bank details are entered with Stripe and never reach our servers)United States
DocuSignElectronic signatures for the client agreements: the onboarding envelope and any agreement an order needs, countersigned by GovSmithThe signers' names, email addresses and IP addresses, the agreement documents and the signing recordUnited States
CloudflareTurnstile bot protection on the invitation pageBrowser signals used to tell people from botsGlobal
Sentry (Functional Software)Error and performance monitoring for the app and its servers, and AI request timings and costs copied from OpenRouterError reports and timings with your user ID, organization ID, page and browser, never your email or name; session replays with all text, inputs and media masked; AI request metadata, never prompts or answersUnited States
Logo.devCompany logos looked up from a website domainCompany website domainsUnited States
SAM.gov, USAspending, Grants.gov, SBIR.gov, acquisition.gov and international tender sources (TED, UK Contracts Finder and Find a Tender, AusTender, CanadaBuys, UNGM)Public procurement, award and entity data for Search, Briefs and company recordsSearch queries and public identifiers we look up (such as a UEI); these sources send us public notices, not your dataVarious government systems
Google and Microsoft (sign-in)Sign in with Google or MicrosoftYour name, email and profile photo at sign-inUnited States
Connected apps: Google Workspace, Microsoft 365, HubSpot, Salesforce, Atlassian, LinkedInOptional connectors a client or a vCM turns on: Google Drive, OneDrive and SharePoint in the Library's Cloud, the others in Settings, Connectors. Other apps (such as Dropbox, Box, Asana, Calendly, DocuSign and GitHub) are connected only on requestOnly the data within the scopes you approve, only while the connector is on; files you import are copied into your LibraryWhere each app keeps your data

Retiring at the cutover

The vendors below served the platform GovSmith replaces and are struck from the table above. Records from that platform stay with them only until they move to GovSmith's own systems; GovSmith's own systems never read or write them, and each is shut down at the cutover gate.

Vendor (struck)What it heldRetires
MongoDB AtlasThe database of the platform GovSmith replaces; its records move to NeonAt the cutover gate
PineconeThe search index over files on the platform GovSmith replaces; search now runs on Neon (pgvector)At the cutover gate
Redis CloudThe cache and message queue of the platform GovSmith replaces; GovSmith runs without oneAt the cutover gate

GovSmith staff

The vCMs assigned to your company and our administrators see your data as section 6 describes. Our staff may be employees or independent contractors; we are responsible for how they handle your data.

Inside your company

Your teammates see the company's Portal. The Owner and Admins can see the team, its roles and quote approvers, and Library activity, and can add or remove people. If you leave the company's team, the content you contributed stays with the company.

Legal, safety and corporate

  • To comply with a law, regulation, subpoena, court order or lawful government request, including export control and sanctions requirements. Where lawful and practical, we notify the affected client company first.
  • To protect the rights, property or safety of GovSmith, our clients or the public, including investigating fraud, abuse or security incidents.
  • In a merger, acquisition, financing or sale of assets, in which case the successor is bound by this Policy for the data transferred, and we will notify you.
  • With your direction or consent, for example when you share a file with your vCM, invite a teammate, or connect a third-party account.

6. Who can see your company's data

  • Your company's own people: everyone your Owner or an Admin invites can use the Portal. The Owner and Admins can also see the team, its roles and quote approvers, and Library activity.
  • Your vCMs: GovSmith staff with an active assignment to your company. When an assignment ends, so does that person's access.
  • View as client: a vCM serving your company may open a read-only preview of your Portal to see exactly what a member of your team sees. The preview changes nothing and sends nothing.
  • GovSmith administrators: the people who run the business can see every client's data for assignments, quality review, billing, security and support. They must use multi-factor authentication.
  • GovSmith AI Workspaces: a staff member's AI workspace that covers your company is shared only with vCMs assigned to your company.
  • Staff-only records, such as internal notes, estimates and the cost of our own work, are not shown in the Portal. You can still ask for any personal data in them (section 10).
  • When your Owner or an Admin removes you, your access ends at once. What you contributed stays with the company.

7. Cookies and browser storage

The Portal uses only the cookies and browser storage it needs to work. It sets no analytics or advertising cookies, and we do not sell or share what any of them record.

NamePurposeTypeLifetime
Refresh token cookie (refreshToken)Keeps you signed in during a session and lets the app renew your accessStrictly necessary; HttpOnly, Secure, SameSite=StrictUntil you close the browser; a session also ends after 60 minutes without activity and after 24 hours at most
CSRF token cookie (csrf-token)Protects forms and API calls from cross-site requests; the app reads it, so it is not HttpOnlyStrictly necessary; Secure, SameSite=Strict24 hours
Access token (session storage)Proves to our servers that you are signed inStrictly necessary15 minutes; removed when the tab closes
Local storage (your profile summary, the CSRF token, the last activity time and a sign-out marker)Draws the app quickly and keeps the 60-minute idle limit and sign-out in step across your tabsStrictly necessaryUntil you sign out or clear it
Local storage (theme, language, layout and unsent drafts)Remembers your preferences and unsent drafts on this deviceFunctionalUntil cleared
Sentry (session storage)Links an error report to the masked replay of the session that led to itError monitoringUntil the tab closes
Cloudflare TurnstileTells people from bots on the invitation pageStrictly necessaryA few minutes
AWS WAF token cookie (aws-waf-token), on govsmith.comTells people from bots on the website's discovery and demo forms; set by a silent check when a form page loadsStrictly necessaryShort-lived; renewed when a form page loads

Our website at govsmith.com uses no analytics or advertising cookies or scripts, and neither does the Portal.

You can clear cookies and browser storage in your browser. Doing so signs you out and resets preferences on that device. Because we do not sell or share personal information, there is nothing to opt out of; we nonetheless honor Global Privacy Control (GPC) signals as an opt-out of any sale or sharing.

8. How we protect information

  • Encryption in transit with TLS, and at rest: our files, secrets and backups with keys held in AWS Key Management Service, and the database with Neon's encryption at rest.
  • Passwords hashed with a modern algorithm; multi-factor authentication required for all GovSmith staff and administrators and available to every client user, with a passkey or an authenticator app; single-use, expiring links for password reset; sessions that end after 60 minutes idle or 24 hours.
  • Client isolation enforced in every query and, as a second wall, by database row-level security; staff access limited to the clients an active assignment names; an automated isolation test on every change.
  • Malware scanning of uploaded files (Amazon GuardDuty), threat detection and audit trails on our cloud account (GuardDuty and AWS CloudTrail), and vulnerability scanning (Amazon Inspector).
  • Audit logs of sign-ins, security events, assignments, approvals and administrative actions; least-privilege access to production, with production credentials kept only in AWS Secrets Manager.
  • Nightly encrypted backups locked against change, restores drilled before launch and every quarter, and a written incident response plan.
  • We self-assess against CMMC Level 1, the basic safeguarding of Federal Contract Information. We are not assessed at CMMC Level 2, we are not FedRAMP authorized, and GovSmith does not handle Controlled Unclassified Information (CUI). Do not share CUI or classified information with us.

No method of transmission or storage is completely secure. If we confirm a breach that affects your personal data or Client Content, we will notify the Owner of the affected client account without undue delay and within 72 hours of confirming it, and individuals and regulators where the law requires.

9. How long we keep information

DataRetention
Account and profileWhile the account is active; an account deletion request is processed within 7 days
Client Content (files, company profile, pursuits, Briefs, Quotes, Deliverables, messages)For the engagement, until you delete it or ask us to; deleted 30 days after the engagement ends
GovSmith AI conversations, workspaces and memories about your companyDeleted with your Client Content, or sooner when our staff delete them
Embeddings and the search indexDeleted with the files they came from
Connector tokensUntil the connector is disconnected, when the token is revoked; files you imported stay in your Library until deleted
Usage and delivery time records24 months, then aggregated
AI request records (model, provider, tokens, cost and timing; never prompts or answers)13 months
Security and audit logs12 months; longer if needed for an investigation
Support requests and bug reports24 months
Discovery inquiries and demo requests sent through govsmith.com24 months, then deleted automatically
Billing records, invoices, and records of accepted Service Agreements and Quotes7 years, as tax and contract law require
BackupsNeon's restore history: up to 30 days. Our nightly encrypted copies: daily copies 35 days, monthly copies 12 months, yearly copies 3 years. They are locked against change, so deleted data leaves them only as each copy expires, and we use them only to recover the Service

We may keep de-identified or aggregated data that cannot reasonably be linked to you or your company for as long as it is useful.

10. Your privacy rights and choices

Wherever you live, you can do the following in the Service:

  • See and update your profile in Settings, Account; manage sign-in security in Settings, Security & login.
  • Download your data in Settings, Data controls.
  • Ask to delete your company's data with Delete all data in Settings, Data controls. The request goes to your vCM, who confirms the scope with you before anything is removed.
  • Request deletion of your own account in Settings, Account. We process it within 7 days.
  • Choose which GovSmith notifications you get, by email or in the app, in Settings, Notifications. Email about your account, security and billing cannot be turned off while the account exists.

Depending on where you live you may also have the right to: access the personal data we hold about you and receive a copy in a portable format; correct inaccurate data; delete your data; restrict or object to certain processing; opt out of targeted advertising, sale, sharing or profiling (we do none of these); withdraw consent where processing is based on consent; and not be discriminated against for exercising your rights.

To exercise a right that the Service does not let you exercise yourself, email legal@govsmith.com with the subject "Privacy request". We verify requests using the email on the account and may ask for more information. We respond within 45 days under U.S. state laws (extendable once by 45 days with notice) and within one month under the GDPR (extendable by two months for complex requests). An authorized agent may submit a request with your written permission. If we deny a request, we explain why, and you may appeal by replying to our decision; we answer appeals within 45 days.

If you are on a client company's team, we may refer your request to that company's Owner, because the company controls its Client Content.

11. Notice for California residents

This section supplements the rest of the Policy for California residents under the California Consumer Privacy Act as amended by the California Privacy Rights Act (CCPA).

Category of personal informationCollectedSourcePurposeDisclosed to
Identifiers (name, email, username, IP address)YesYou; your company; sign-in providersProvide the Service, security, billing, supportService providers in section 5
Commercial information (Service Agreement, Quotes, Orders, invoices, payments)YesYou; StripeBilling; providing the ServiceStripe
Professional information (job title, company, role)YesYou; your company; public registriesProvide the ServiceService providers in section 5
Internet activity (usage, device, logs, masked session replays)YesAutomaticallyProvide the Service, reliability, securityHosting and monitoring providers
Geolocation (approximate, from IP; addresses you give us)YesAutomatically; youSecurity; place lookups for your recordsHosting providers (Amazon Location Service)
Sensitive personal information (account credentials)YesYouSign-in onlyNone
InferencesLimited (win estimates and fit scores about your company's opportunities; GovSmith AI memories our staff keep)Our work for your companyProvide the ServiceAI providers in section 5, when our staff use them in the work

We do not sell personal information and do not share it for cross-context behavioral advertising. We do not use or disclose sensitive personal information for purposes other than those permitted by the CCPA. We honor Global Privacy Control signals. California residents have the rights to know, delete, correct, and to limit the use of sensitive personal information, and the right not to be discriminated against; section 10 explains how to exercise them.

12. Notice for residents of other U.S. states

Residents of Virginia, Colorado, Connecticut, Utah, Texas, Oregon, Montana, Iowa, Indiana, Tennessee, Delaware, New Jersey, New Hampshire, Kentucky, Nebraska, Maryland, Minnesota, Rhode Island and other states with comprehensive privacy laws have the rights listed in section 10, including the right to appeal a denied request. We do not sell personal data, engage in targeted advertising, or profile you in furtherance of decisions that produce legal or similarly significant effects.

Connecticut residents: we do not collect, use or sell personal data for the purpose of training large language models. Where a state law requires it, we treat a browser or device opt-out preference signal such as Global Privacy Control as a valid opt-out.

13. International data transfers

The Service is hosted in the United States (AWS us-east-1), and our subprocessors are mainly in the United States; some AI providers in section 5 may answer a request from their own regions. If you use the Service from outside the U.S., including from the European Economic Area, the United Kingdom or Switzerland, your information is transferred to and processed in the U.S., where privacy laws may differ from those in your country.

For transfers from the EEA, UK and Switzerland we rely on the European Commission's Standard Contractual Clauses and the UK Addendum with our clients and subprocessors, and on the necessity of the transfer to perform our contract with you. Client companies that need signed clauses should email legal@govsmith.com. We have not appointed an EU or UK representative; contact us directly.

14. Children

The Service is for business use and is not directed to anyone under 18. We do not knowingly collect personal information from children. If you believe a child has given us personal information, contact us and we will delete it.

15. Government contractor considerations

  • GovSmith does not handle Controlled Unclassified Information (CUI) or classified information. Do not upload or share either. If either reaches us, we will tell you and remove it, and we may be required to report it.
  • You decide whether Client Content includes technical data controlled under ITAR or the EAR or other regulated data, and you share it only as our Terms of Service allow.
  • Public procurement data we use (SAM.gov, USAspending, Grants.gov, SBIR.gov and international sources) is public information and is not personal data we collect from you, although it may name government officials and business contacts.
  • We may be required to preserve or disclose data in response to lawful process, including subpoenas related to your government contracts; we notify you where the law allows.
  • We do not offer a HIPAA Business Associate Agreement. Do not share protected health information with us.

16. Automated decision-making and profiling

Our staff use AI and simple scoring tools to draft, summarize, classify and recommend, for example how well an opportunity fits your company or a suggested win estimate. A person at GovSmith reviews these before they reach you, and you decide what to do with them. We do not use automated decision-making to make decisions that produce legal or similarly significant effects about you, such as denying you access, pricing you differently, or evaluating your employment.

18. Marketing and product communications

We send email about your account, security events, team invitations, your engagement (Briefs, Quotes, Orders, messages and meetings), billing and support; you cannot turn off email about your account, security and billing while you have an account. You choose which other notifications you get in Settings, Notifications. Any news or announcement email we send carries an unsubscribe link. We do not send SMS marketing.

19. Changes to this Policy

We will post any changes here and update the effective date. For material changes we will notify the Owner of each client account by email or a notice in the Portal at least 30 days before they take effect. Earlier versions are available on request.

20. Contact us

Questions, requests and complaints about privacy:

  • Blacksmith AI LLC DBA "GovSmith"
  • 1611 N. Broadway Ave.
  • Oklahoma City, OK 73103
  • United States
  • Legal and privacy requests: legal@govsmith.com
  • Support: support@govsmith.com
  • Web: https://govsmith.com

If you are in the EEA or UK and are not satisfied with our response, you may lodge a complaint with your local data protection authority. Previous version: October 21, 2026. This version is effective October 31, 2026.