TOGO Responsible AI & Public Indexing Policy

This policy applies to AI-assisted activity connected with TOGO's public website, product discovery, multilingual content, customer-support drafting, internal search, and aggregated website analytics. It describes our present operating approach and may be updated as technologies, law, and business practices change.

• Human responsibilityNamed teams own decisions and published outcomes.
• Proportionate useHigher-risk claims receive stronger checks or no AI use.
• Data minimizationOnly information reasonably needed for an approved task should be used.
• Traceable sourcesMaterial facts should be checked against current records and documents.
• Clear limitationsWe avoid presenting predictions or generated text as verified facts.
• Continuous improvementFeedback, errors, and changing requirements inform our controls.

Approved tools may support limited, reviewable tasks such as:

• Editorial supportOutlines, summaries, readability improvements, and metadata drafts.
• Translation supportFirst-pass localization that is reviewed for product and market context.
• Search & discoveryHelping visitors locate public pages about bagasse, paper, bamboo, wood, or PLA products.
• Analytics supportIdentifying aggregate content trends without treating a model inference as customer intent.

We do not represent that every page or sentence is AI-assisted. Workflows vary by task, sensitivity, source quality, and the availability of qualified reviewers.

TOGO personnel review AI-assisted material in proportion to its impact. Editors check public content; sales and operations teams confirm commercial details; and appropriate technical or quality personnel review product-specific claims when supporting evidence is required.

1. 1
   Source verification

   Begin with current, authorized information.

2. 2
   Assisted draft

   Use tools only within the approved task boundary.

3. 3
   Human review

   Check facts, context, tone, and unsupported implications.

4. 4
   Official confirmation

   Issue formal business or production information through authorized people.

Public website content is different from non-public information. Customer names, contact details, drawings, samples, target pricing, order data, contracts, and unpublished product plans are not made public merely because they are provided to TOGO.

AI systems can produce incomplete, outdated, or incorrect output. Before publishing material facts, we aim to compare them with relevant product data, test reports, internal records, supplier information, or responsible subject-matter review.

• Material composition and intended-use statements
• Dimensions, capacities, packaging, and customization options
• Food-contact, compostability, recycling, and environmental claims
• Certification scope, dates, product coverage, and destination-market relevance

If you believe a page contains an error, contact us with the URL and supporting context. We will assess appropriate correction, clarification, or removal.

AI does not independently control TOGO production, release finished goods, approve a sample, or certify conformity. Product outcomes depend on confirmed material, tooling, tolerances, printing, packaging, production conditions, inspection criteria, and agreed order documents.

• Materials varyBagasse, paper, bamboo, wood, and PLA have different performance and end-of-life considerations.
• Use cases varyTemperature, food type, contact time, storage, and local infrastructure affect suitability.
• Markets varyClaims and compliance requirements may differ by country or region.
• Orders are specificOnly accepted samples and written order documents define the agreed supply.

We support responsible discovery of genuinely public pages. Public visibility, however, does not erase intellectual-property rights, contractual restrictions, technical access controls, or privacy obligations.

Search engines, research crawlers, AI discovery services, and other automated systems—including systems operated by Google, Microsoft Bing, OpenAI, Anthropic, Perplexity, and similar providers—may encounter our public website. Naming a provider describes a category of discovery technology and does not imply partnership, approval, or endorsement.

• Follow robots.txt, page-level directives, authentication, and other technical controls.
• Use reasonable request rates and do not degrade availability or security.
• Do not evade blocks, scrape restricted areas, or attempt to reconstruct private information.
• Respect copyright, database rights, trademarks, terms, privacy, and applicable law.

We may change automated-access permissions, restrict particular agents, or take technical and legal measures where activity creates risk, excessive load, misuse, or non-compliance.

Unless TOGO gives prior written authorization, automated systems must not:

• MisrepresentPresent generated or altered output as an official TOGO quotation, certificate, test report, or endorsement.
• ImpersonateCreate deceptive accounts, communications, products, or services using our identity or marks.
• Exploit accessBypass security, authentication, rate controls, or instructions for automated access.
• Reuse unlawfullyBuild unauthorized replicas, datasets, or commercial offerings from protected TOGO content.
• Infer private factsCombine data to identify customers, pricing, orders, suppliers, or other non-public relationships.
• Create harmUse our content for fraud, unsafe product claims, discrimination, malware, or unlawful activity.

Customers may ask whether AI assistance was used in preparing a material communication, request human review, identify an accuracy or privacy concern, and seek official written confirmation for commercial or technical matters. We will evaluate requests in light of available records, confidentiality, law, and operational feasibility.

Responsibility for AI-assisted activity is distributed across the relevant business, editorial, technical, privacy, quality, and management functions. Controls may include tool approval, access limits, source requirements, reviewer assignment, incident escalation, correction processes, and periodic reassessment.

This policy does not create a warranty or expand contractual obligations. We may revise it to reflect operational learning, technology changes, legal requirements, or updated public indexing choices. The date at the top identifies the latest published version.