Brand Safety and Sensitive Content Policy
Version: 0.1-draft Effective date: not yet in force (draft)
Draft pending review by a qualified Romanian lawyer before use.
1. Purpose and scope
1.1 This policy sets out how the 1ads Platform, operated by ALTTAB PROFIT SRL, a limited liability company (the "Company"), treats sensitive topics. It defines the minimum safeguards that apply to every Customer (the "platform floor") and the options Customers have to make them stricter.
1.2 It applies to every AI Feature that suggests, generates or schedules content, in particular Trend Radar concepts, AI-generated copy and images, ideas derived from Web Intelligence, and the scheduling and publishing tools, for both organic posts and paid ads.
1.3 This policy forms part of the Terms (terms-of-service.md) and complements: (a) the Acceptable Use Policy (acceptable-use-policy.md), which sets out what Customers may not do; and (b) the AI Transparency Notice (ai-transparency-notice.md), which explains the AI Features and AI labelling. If this policy and the Acceptable Use Policy set different rules for the same situation, the stricter rule applies. The AI Transparency Notice is informational (Section 1.3 of that notice).
2. Principles
2.1 A floor, not a ceiling. The platform floor is the minimum. Customers can add rules but cannot remove or relax them.
2.2 When in doubt, escalate. If the Platform cannot classify a topic with confidence, it treats it as the more sensitive category.
2.3 Purpose before promotion. On sensitive topics, the default is to say something useful or respectful, not to sell.
2.4 People decide. The Platform classifies and proposes. Authorised people approve. Nothing on sensitive topics is ever published automatically.
2.5 Transparency. Each trend and concept shows its classification and the main reasons for it, so users can understand and challenge it.
3. Classification levels
3.1 Every trend, concept and draft is assigned one of the following levels:
| Level | What the Platform does |
|---|---|
| Blocked | No concepts or drafts are generated. Trend Radar discards the trend. The Platform does not schedule or publish content on the topic. |
| Gated | Concepts may be generated only in the "purpose-only" posture (section 5.2), unless an exception is recorded (section 5.3). Approval by a Senior Approver is required in addition to the normal approval workflow. Never auto-approved, never auto-published. |
| Standard | Normal approval workflow and the Customer's brand rules apply. |
3.2 Senior Approver means an Authorized User to whom the Customer has assigned this role in the Platform. The Customer should assign it only to people with the authority and experience to judge reputational and legal risk (for example a head of communications or marketing, or a person designated by the agency's client).
4. Blocked categories (platform floor)
4.1 The following topics are Blocked for all Customers:
| # | Category | Examples |
|---|---|---|
| B1 | Commercial exploitation of tragedies | Promotions, offers, product placement or trend-jacking linked to natural disasters, accidents, terrorist attacks, mass shootings, armed conflicts, epidemics, deaths, missing persons or periods of national mourning |
| B2 | Violence and extremism | Graphic violence, gore, glorification of terrorism or violent extremism |
| B3 | Hate and harassment | Attacks on people based on protected characteristics, dehumanising content, bullying, doxxing |
| B4 | Sexual content | Explicit sexual content; non-consensual intimate imagery; any sexualisation of minors (child sexual abuse material is reported to the competent authorities) |
| B5 | Self-harm | Promotion or trivialisation of suicide, self-harm or eating disorders |
| B6 | Illegal goods and activities | Illegal drugs, weapons and explosives trade, counterfeits, human trafficking, hacking services |
| B7 | Harmful misinformation | Health misinformation, conspiracy theories, manipulated news, election misinformation |
| B8 | Elections and political actors | Content that promotes or opposes a political party, candidate, elected official in a political capacity or referendum option; content liable and designed to influence the outcome of an election or referendum, voting behaviour or a legislative or regulatory process; political advertising within the meaning of Regulation (EU) 2024/900 (see section 10) |
| B9 | Deceptive synthetic media | Deepfakes of real persons without consent, impersonation, fake endorsements or testimonials |
| B10 | Exploitation of vulnerable people | Mocking victims, exploiting people's distress, fear or financial hardship, or the vulnerability of minors |
| B11 | Age-restricted products aimed at minors | Any content for alcohol, gambling, tobacco or nicotine products, or adult products that targets or appeals mainly to minors |
4.2 Category B1 blocks the commercial use of these events. Non-promotional messages about them (condolences, safety information, verified relief information) are handled under gated category G3 (section 5.1).
5. Gated categories
5.1 The following topics are Gated for all Customers:
| # | Category | Examples |
|---|---|---|
| G1 | Social movements and causes | #MeToo, gender equality, LGBTQ+ rights and Pride, racial justice (for example #BlackLivesMatter), climate activism, migrants and refugees, disability rights, mental-health awareness |
| G2 | Commemorations and remembrance | Holocaust Remembrance Day, anniversaries of tragedies (for example the Colectiv fire), the 1989 Revolution, Heroes' Day, international days for victims |
| G3 | Solidarity after tragedies and disasters | Condolences, safety information, verified relief or donation information, service disruption notices |
| G4 | Health and medicine | Diseases, treatments, pandemics, mental health, body image |
| G5 | Religion | Religious topics, figures, practices and controversies (ordinary seasonal greetings such as Christmas or Easter are Standard, with a tone check) |
| G6 | Identity and history | Ethnicity, nationality, national minorities, relations between communities, national symbols, sensitive historical events |
| G7 | Real people in the news | Celebrities and public figures, deaths, scandals, legal proceedings |
| G8 | Crime and justice | Crimes, investigations, trials, law enforcement |
| G9 | Economic distress | Layoffs, inflation, bankruptcies, bank failures, strikes |
| G10 | Children and families | Content featuring or addressed to minors |
| G11 | Public policy and civic topics | Informational content about legislation, regulation and public debates, and non-partisan civic messages that do not seek to influence an election, a referendum, voting behaviour or a legislative or regulatory process (organic only; see section 10) |
5.2 Purpose-only posture. By default, a concept on a gated topic must: (a) contain no product, price, offer, discount or call to buy; (b) be connected to a genuine commitment of the brand that the Customer has documented in the brand profile (for example a policy, a programme, a partnership or a track record); (c) not use a movement's slogans, symbols or hashtags for promotion; (d) not speak on behalf of survivors or affected communities, and prefer amplifying credible organisations and voices; (e) use content warnings where the topic may distress people; (f) be organic only, with no paid promotion to EU audiences, and not be liable and designed to influence the outcome of an election or referendum, voting behaviour or a legislative or regulatory process (see section 10).
5.3 Exceptions to the purpose-only posture. For categories G1, G4, G5, G6, G10 and G11, a Senior Approver may approve a specific concept that departs from section 5.2(a) where there is a documented justification (for example a long-standing cause partnership with a documented donation programme). The exception and its reason are recorded in the audit log. No exceptions are possible for G2, G3, G7, G8 and G9; for these categories the purpose-only posture always applies.
6. Restricted verticals
6.1 Gambling, alcohol, tobacco and nicotine products, medicines and medical devices, food supplements, financial and investment products, crypto-assets, weapons, adult products, dating services and cosmetic medical procedures are restricted verticals.
6.2 Content for a restricted vertical is available only after the Customer enables the vertical in the brand profile and confirms that it holds the necessary licences and will comply with the applicable advertising rules and age restrictions. The Platform then adds the vertical's specific checks (for example mandatory warnings and age targeting) to the approval workflow. [TBC: list of checks per vertical.]
7. Brand-level configuration (stricter only)
7.1 For each brand, the Customer can: (a) add blocked topics, keywords, hashtags, people, competitors and "never mention" lists; (b) move any gated category to blocked; (c) require additional approvers or approval steps; (d) restrict the markets and languages for which Trend Radar proposes concepts, or turn Trend Radar off for a brand or market; (e) set tone-of-voice rules and other brand guidelines.
7.2 The Customer cannot: (a) unblock a blocked category; (b) remove the Senior Approver requirement for gated categories; (c) remove AI labels or markings applied by the Platform; (d) enable automatic publication of Trend Radar concepts or of content on gated topics.
7.3 Enterprise customers may discuss additional controls with us. The platform floor applies to them as well.
8. How the gate works
8.1 Sources. Trend Radar uses only public sources: Google Trends (via SerpApi), YouTube, Wikipedia, GDELT, news APIs, Bluesky and Mastodon. Web Intelligence reads public web pages only, respecting robots.txt and text-and-data-mining opt-outs.
8.2 Steps. Each trend passes through the following steps: (a) collection and de-duplication; (b) classification using rule and keyword lists and AI classifiers, in several languages; (c) assignment of a level (Blocked, Gated or Standard), applying section 2.2 where the classification is uncertain; (d) matching against the brand profile and its exclusions; (e) concept generation within the limits of the assigned level and posture; (f) a second check of each generated concept, copy and image; (g) human approval, including a Senior Approver for gated topics; (h) a final check shortly before publication, because the context can change between approval and publication (for example if a tragedy occurs in the meantime). If the level has changed, the publication is held for re-approval.
8.3 Errors. Classification can be wrong: it may miss sensitive content or block harmless content. Users can report a misclassification from the Platform [TBC] or at support@1ads.net. The final responsibility for what is published remains with the Customer.
9. Crisis mode
9.1 Brand crisis mode (activated by the Customer). A Senior Approver or an administrator of the Customer can activate crisis mode for a brand, for example during a product recall, a security incident or a reputational crisis. While it is active: (a) all scheduled publications for that brand are held and need re-approval; (b) Trend Radar stops proposing concepts for the brand; (c) Performance Guardian keeps monitoring and flags proposals that would increase spend or launch new campaigns, so they can be reviewed in the light of the crisis; (d) the AI chat shows a crisis-mode notice. Crisis mode is deactivated by a Senior Approver or an administrator. Activation and deactivation are recorded in the audit log.
9.2 Market crisis mode (activated by the Company). When a major event affects a market (for example a major tragedy, disaster, terrorist attack, escalation of an armed conflict or a period of national mourning), the Company may activate crisis mode for that market or language. While it is active: (a) Trend Radar stops proposing concepts related to the event and applies stricter classification in that market; (b) publications scheduled in the affected market for the next 48 hours are held until an Authorized User confirms them again; (c) we notify affected Customers in the Platform and by email.
9.3 National mourning in Romania. A day of national mourning declared by the Romanian authorities activates market crisis mode for Romania by default.
9.4 Ads on Third-Party Platforms. Crisis mode does not pause ads running on Third-Party Platforms. The Platform may generate proposals to pause or adjust campaigns, which run only after an authorised person approves them.
9.5 End of crisis mode. The Company reviews market crisis mode at least daily and ends it when the reason no longer applies.
10. Political advertising in the EU
10.1 Regulation (EU) 2024/900 on the transparency and targeting of political advertising (TTPA) has applied since 10 October 2025. The Platform does not support political advertising in the EU. Meta stopped serving political, electoral and social-issue ads in the EU, and Google stopped serving political ads in the EU, from October 2025.
10.2 Consequently: (a) election and political-actor topics are Blocked (B8); (b) public policy, civic and social-issue topics (G1, G11) are allowed only as organic content that is not liable and designed to influence the outcome of an election or referendum, voting behaviour or a legislative or regulatory process, and never as paid ads aimed at EU audiences. Content that is liable and designed to have that effect is political advertising under Article 3(2) of Regulation (EU) 2024/900 even when it is not paid for by the brand, where it is produced for remuneration (for example by an agency) or through in-house activities or as part of a campaign. It is Blocked (B8); (c) paid promotion of an organic post on these topics, or its production as part of a campaign for a political actor, may turn it into political advertising and is not supported.
10.3 Outside the EU, political advertising is not supported during the closed access programme either, and no Order Form can make an exception (acceptable-use-policy.md, section 3.5).
11. Responsibility and limits
11.1 The platform floor reduces risk but cannot remove it. The Customer is responsible for the content it approves and publishes, and for complying with the law and Third-Party Platform policies.
11.2 Brand safety in the sense of ad placement (where ads appear next to other content on a Third-Party Platform) is controlled by each Third-Party Platform's inventory and suitability settings. The Customer should configure those settings in its ad accounts. The Platform may recommend settings but does not control the placement of ads by Third-Party Platforms.
12. Review and updates
12.1 We review this policy at least every quarter and after major events. Changes that make the platform floor stricter may apply immediately (Section 26.3 of the Terms). Other material changes are notified in advance, as described in Section 26.2 of the Terms.
12.2 Suggestions and reports: support@1ads.net. Reports of illegal content: support@1ads.net.