Moderation Policy

Version 1.2 · Toke AI Ltd

Effective: 1 September 2026

Owner: Byron Coop, Chief Technology Officer

Accountable individual: Chief Executive Officer

Applies to: dispensary reviews

Review: annually, and before any significant change to the service

Published at toke.ai/policies alongside our Privacy Policy, Terms of Service and Cookie Policy.

Our commitment

A plain-language summary of what follows. Where this summary and the detailed sections differ, the detailed sections govern.

Toke is a community built around a controlled substance. That places a higher duty on us than on a general-purpose platform, and we accept it.

We commit to keeping unlawful supply off the platform, to protecting people from harassment and harm, to keeping minors out, and to being clear and fair when we act on someone's content or account.

What we do

What is not allowed

In summary. Set out in full at section 4.

• Offering, requesting or arranging supply of cannabis
• Encouraging anyone to use cannabis
• Medical or dosing advice, or claims that a product treats a condition
• Anything involving or aimed at under-18s
• Guidance on growing, extraction or processing
• Harassment, hate speech, threats or publishing someone's private information
• Reviews by businesses of their own operations or their competitors'

How quickly we respond

Our moderation team works 09:00 to 17:00 in each of our markets, and another market picks up where it can. Resolution times run during those hours. A report that arrives when every market is closed is dealt with first thing the next working day, and the most serious reports go to the front of the queue.

Age

Toke is for adults only. You confirm your age when you create an account, and we check an age signal from the app store when you first open the app. If that signal says you are under 18, the account is not activated.

Being fair

• We tell you what we removed and which rule it broke
• You can appeal by replying to the message we send you, or by writing to support@toke.ai
• A person reviews it, not an automated system
• If we get it wrong, we restore the content and reverse the action
• Users in the EU can also refer disputes to a certified out-of-court body

Who is responsible

Our Chief Executive Officer is accountable for moderation and for our duties under the Online Safety Act.

Day-to-day moderation is handled by our customer support team in each market. Appeals are handled by our Head of Support, escalating to our Chief Technology Officer and then our Chief Executive Officer where needed.

As the platform grows we will move moderation to a dedicated role. The thresholds at which we do that are at section 10.3, and we review them against actual performance rather than by judgement.

1. Purpose and scope

This policy sets out what may and may not be posted on Toke, how content is reviewed, what happens when a rule is broken, and how a user can challenge a decision.

Toke is a cannabis information platform. Cannabis is a controlled substance in every territory in which Toke operates, and its legal status differs materially between them. That combination places obligations on us that a general-purpose community would not carry, and this policy is written to meet them.

1.1 Surfaces covered

Some dispensaries require a user to be at the venue before a review can be submitted. That is an eligibility condition rather than a moderation decision, and it is described in section 4.7 of our Privacy Policy. Reviews already submitted are unaffected by it.

Not in this release. Direct messaging, friend connections, strain reviews, image attachments and community forums are not part of this release. Each requires this policy to be reviewed, and a fresh illegal content risk assessment, before it ships.

1.2 Who this applies to

Every account holder, including business users representing dispensaries or brands. Business users carry the additional obligations at section 4.6.

2. Regulatory mapping

Toke is a micro enterprise for the purposes of Article 19 of the Digital Services Act, and the online platform obligations in Section 3 of Chapter III do not apply. The obligations above are not within that section and apply regardless.

3. Definitions

Content — anything an account holder submits or publishes, including review free text, structured ratings, display names and profile information.

Illegal content — content amounting to a criminal offence in a territory where the service is available, whether or not a prosecution follows.

Proselytism — advocacy or encouragement of drug use directed at another person, distinct from arranging supply. Actionable in Italy including in private communication.

Report — a notification that content may breach this policy or the law.

Moderator — a person authorised to review content and apply enforcement actions under section 7.

4. Prohibited content

The prohibitions below are cumulative. Content may breach more than one.

4.1 Supply and sourcing

Unlicensed supply of cannabis is a criminal offence in every territory in which Toke operates. This is the highest-severity category on the platform.

• Offering to supply, sell, share or gift cannabis or cannabis products
• Soliciting supply from another user
• Arranging a meeting, exchange or transaction for the purpose of supply
• Sharing contact details, handles or off-platform destinations for the purpose of arranging supply
• Diverting, reselling or sharing prescribed medication
• Directing users to unlicensed sources, including coded, indirect or euphemistic references

Permitted: a user describes their experience of a dispensary and mentions obtaining a product there. Not permitted: a user says they have some available, invites others to contact them, or asks where to get something outside a licensed channel.

4.2 Proselytism

A separate prohibition from section 4.1. Article 82 of Italian DPR 309/1990 reaches proselytism for drug use carried out privately as well as publicly.

• Encouraging or pressuring another person to use cannabis
• Encouraging use by a person who has indicated they do not use, are abstaining, or are in recovery
• Content framing use as something the reader should take up

Permitted: describing personal experience, including positive experience. Not permitted: telling another user they should try it, or that abstaining is a mistake.

4.3 Health and medical

• Medical advice, including dosing, titration or how to combine products
• Advising another user to change, reduce or stop prescribed medication
• Claiming that any product or strain cures, treats or prevents a named condition
• Discouraging a user from seeking professional medical care
• Presenting personal experience as clinical evidence

Users will write condition-level claims in free text regardless of how the review fields are structured. Removing them is required, because publishing them would present a product as having properties for treating disease.

Nothing on the platform is a substitute for advice from a qualified healthcare professional, and our Terms of Service say so.

4.4 Minors

• Any content directed at, involving or depicting a person under 18
• Content promoting or normalising consumption to minors

Content in this category is escalated immediately under section 9 and is not subject to the graduated enforcement ladder.

4.5 Cultivation

Toke provides no cultivation tools, guidance or software. User-generated cultivation content is prohibited outright rather than geo-gated, because legality differs between territories — lawful within limits in Germany, tolerated privately in Spain, a criminal production offence in the UK — and a single rule is enforceable without reference to any user's location.

• Instructions or guidance on growing, propagating or harvesting
• Guidance on extraction, concentration or processing
• Sourcing of seeds, clones or cultivation equipment
• Depictions of cultivation in progress

4.6 Business users

• Owners, employees and volunteers of a dispensary or brand may not review their own operation or a competitor's
• Incentivising, soliciting or compensating reviews
• Submitting or coordinating reviews from multiple accounts

4.7 General conduct

• Harassment, bullying, threats or intimidation
• Hate speech or discriminatory content targeting a protected characteristic
• Publishing another person's private information
• Impersonating another person, business or Toke itself
• Sexual content, or sexualised commentary about an identifiable person
• Spam, unsolicited advertising or automated posting
• Content infringing another party's intellectual property

4.8 Oncology claims

The Cancer Act 1939 prohibits advertisements offering to treat cancer, prescribing a remedy for it, or offering advice on its treatment. It reaches advice rather than promotion alone, and is therefore broader than medicines advertising law. Oncology treatment claims are removed on sight in all territories.

5. Detection and moderation

Content is reviewed against this policy through three routes.

Content assessed as breaching section 4 is removed under section 7, and the account holder is notified under section 7.4.

6. Reporting and notice-and-action

Any account holder may report content from within the app. Reports are also accepted at support@toke.ai from anyone, satisfying Article 16 of the Digital Services Act, which requires a mechanism that is easy to access, user-friendly and allows electronic submission.

Content is not visible without an account, so in-product reporting is available to account holders only. The published address exists so that a person without an account who becomes aware of content can still notify us.

6.1 What a report captures

• The content complained of and its location
• The category of concern
• An explanation from the reporter
• Contact details, unless the report concerns an offence where anonymity is appropriate

6.2 Response standards

Reports are handled through our internal ticketing system, which is operated by staff through our administration portal rather than within the app, with service levels applied by category. Moderation operates 09:00 to 17:00 in each market.

Acknowledgement is automated and immediate in every case. Resolution targets run during coverage hours, with cross-regional pickup where another market is open.

A report received outside coverage hours in every market is actioned at the start of the next working period, and is placed at the head of the queue where it falls in the top two categories.

7. Enforcement

7.1 Principles

• Proportionate to severity, intent and history
• Consistent between users in comparable circumstances
• Reasoned, with the reason given to the affected user
• Appealable under section 8

7.2 Enforcement ladder

Content falling under section 4.4 bypasses the ladder and results in immediate removal, immediate account suspension and escalation under section 9.

7.3 Blocking

Any account holder may block another. Blocking hides the blocked user's reviews and profile content from the blocker, is available without submitting a report, and is not disclosed to the blocked user.

7.4 Statement of reasons

Every enforcement action is accompanied by a statement to the affected user, as required by Article 17 of the Digital Services Act. It sets out the action taken, the content concerned, the provision of this policy relied upon, whether automated means were used, and how to appeal.

8. Appeals

• Any enforcement action may be appealed within 30 days
• Appeals are submitted by replying to the notification, or to support@toke.ai
• Appeals are decided by a person, not by automated means, in accordance with Article 20 of the Digital Services Act
• Appeals are decided by the Head of Support, escalating to the Chief Technology Officer and then the Chief Executive Officer where the original decision was taken at that level or an alternative view is needed
• Decisions are issued within 14 days with reasons
• Where an appeal succeeds, content is restored and any sanction reversed

Users in the European Union may also refer a dispute to a certified out-of-court dispute settlement body under Article 21 of the Digital Services Act. This does not affect the right to bring court proceedings.

9. Illegal content and law enforcement

9.1 Escalation

Content assessed as illegal is removed and recorded regardless of whether it is reported to authorities. Content within section 4.4, and credible threats to life, are escalated immediately to the accountable individual.

9.2 Disclosure

• Law enforcement requests are directed to the point of contact at section 10.2
• Requests are assessed for lawful basis and proportionality before any disclosure
• Disclosures are logged, including what was disclosed, to whom, and on what basis

9.3 Cross-border requests

Toke operates in territories where cannabis use carries different legal consequences, and users travel between them.

Article 48 of the UK and EU GDPR provides that a judgment or decision of a third-country authority requiring disclosure of personal data is recognisable or enforceable only where based on an international agreement, such as a mutual legal assistance treaty. A direct request from a foreign authority is therefore not, by itself, a lawful basis for disclosure, and complying with one would be a breach in its own right.

• Requests are accepted only through a UK court order, or through a mutual legal assistance treaty or other recognised cross-border instrument
• A direct request from a foreign authority is acknowledged, recorded, and declined pending a valid route, with the correct route identified to the requester
• Requests concerning supply to others, or harm to others, receive cooperation within those channels
• Requests concerning a user's own personal consumption are resisted and require a court order. Where the request originates from a territory in which personal use is criminalised, Toke will consider challenging it
• The affected user is notified unless notification is legally prohibited
• Every request and disclosure is logged, and aggregate figures maintained internally

10. Governance

10.1 Contact

Reports, appeals and general moderation correspondence: support@toke.ai. This address is published at toke.ai/policies and is reachable without an account. It is also our point of contact for recipients of the service under Article 12 of the Digital Services Act.

A contact form is available at toke.ai/contact-us as an additional route.

10.2 Point of contact for authorities

Single electronic point of contact for member state authorities, the European Commission and the European Board for Digital Services, as required by Article 11 of the Digital Services Act: support@toke.ai. Language of communication: English.

10.3 Accountability and resourcing

The Chief Executive Officer is the accountable individual for moderation outcomes and for the illegal content duties under the Online Safety Act. The role carries authority to allocate resource, and is the same individual named in our Online Safety Act risk assessment.

Scaling triggers. Moderation becomes a dedicated function rather than part of an existing role when:

The service level trigger is self-calibrating: it fires whether the cause is volume, case complexity or staff absence.

We keep the resourcing model under review against actual out-of-hours report volume. If material volume arises outside coverage hours, the position is reassessed.

10.4 Moderator training

• Induction covering this policy, the enforcement ladder and escalation routes
• Specific training on sections 4.1 and 4.2, including evasion patterns
• Refresher training at least annually and on material policy change
• Wellbeing support for moderators exposed to distressing content

10.5 Related documents

All three policies are published at toke.ai/policies

10.6 Review

Reviewed at least annually, and before any significant change to the service. Introducing direct messaging, friend connections, strain reviews, image attachments, communities, the medical lane or commerce each constitutes a significant change and requires review before release, together with a fresh illegal content risk assessment.