OriginalVoices Question & Content Policy

OriginalVoices lets you ask questions to real people's Digital Twins and receive insights instantly. To keep the platform safe, respectful, and compliant, all submitted content is automatically reviewed against our content policy. Questions that violate the policy are blocked before they reach any Twin.

Our guiding principle: if a question could reasonably appear in a professional market research survey, it's likely allowed.

Allowed Questions & Topics

You can ask about a wide range of topics, including:

• Consumer behaviour, habits, and routines
• Shopping, brand preferences, and product feedback
• Entertainment, media, and streaming choices
• Technology use and feature preferences
• Travel, food, and lifestyle
• Workplace experiences and hybrid working
• Health experiences and wellbeing (general sentiment)
• Political opinions, policy attitudes, and voting behaviour
• Demographic targeting for audience research
• Sensitive but research-appropriate topics (e.g. drug use, mental health, addiction)

Examples of allowed questions

• "How important is convenience when choosing food delivery apps?"
• "What features matter most when buying headphones?"
• "How do you feel about hybrid working?"
• "What streaming services do you currently subscribe to?"
• "How often do you exercise each week?"
• "Have you considered or are you currently using weight loss medication like Ozempic?"
• "Are you or have you ever taken anti-depressants?"
• "How has your mental health been over the past year?"
• "Have you ever taken recreational drugs?"
• "Have you experienced addiction or substance dependency?"

Blocked Topics

The following categories of content are not permitted:

Sexual content

Explicit, implicit, suggestive, or intimate preferences and fantasies.

❌ "How often do you watch porn?" ❌ "What are your intimate preferences in a relationship?"

Hate and discrimination

Content that attacks, demeans, or promotes hostility toward protected groups (race, religion, gender, disability, orientation, nationality, etc.).

Standard demographic targeting for audience research is allowed — this is normal market research segmentation, not discrimination.

✅ "Women aged 35-55 in London" ✅ "Black consumers aged 18-24" ✅ "How do Muslim parents feel about school meals?" ✅ "What do LGBTQ+ millennials look for in a brand?" ✅ "How do disabled users navigate this app?"

❌ "Which ethnic group do you trust the least?" ❌ "Should immigrants have fewer rights?" ❌ "Why are [group] worse at [thing]?" ❌ "Rank these races by intelligence"

Violence and harm

Self-harm, threats, abuse, and weapons content.

Extremism

Content related to terror groups, radicalisation, or ideological recruitment.

Political manipulation

Using Twins to generate persuasion tactics, campaign messaging, or voter manipulation strategies. Testing reactions to political messaging, policies, ads, and candidates is allowed — this is standard opinion research.

✅ "How familiar are you with the debt ceiling in the US?" ✅ "Which political party do you most identify with?" ✅ "Do you support or oppose the Republican party's tax plan?" ✅ "Here's candidate X's ad — how does it make you feel?" ✅ "Does this policy message resonate with you?" ✅ "What concerns do you have about this immigration proposal?" ✅ "How would you react if the government announced this policy?" ✅ "Do you approve of the current government's performance?" ✅ "How important is immigration policy when deciding your vote?" ✅ "How concerned are you about the rising cost of living?"

❌ "Write a message that would convince young voters to support candidate X" ❌ "What emotional triggers would shift undecided voters?" ❌ "How could we frame this policy to win over conservatives?" ❌ "Generate campaign slogans for the Labour party"

Illegal activity instructions

Instructions on how to commit crimes or justification of crimes. Research questions about behaviours and experiences are allowed (e.g. drug use, tax compliance attitudes).

❌ "How did you get away with not paying taxes?"

Professional advice requests

Specific medical diagnosis/prescriptions, legal strategy, or financial instructions. General health, financial sentiment, and experience questions are fine.

✅ "How do you feel about the cost of private healthcare?" ✅ "Have you ever struggled to get a GP appointment?"

❌ "What medication should I take for anxiety?" ❌ "Is it worth investing in crypto right now?"

Minors in sensitive contexts

Questions involving minors in sensitive situations — unless framed for parents or teachers.

Identifying personal information (PII)

Questions designed to elicit a Twin's real name, address, phone number, email, government ID, employer name, or other information that could identify the individual behind a Twin. Standard demographic questions (age range, gender, city/region) are allowed.

✅ "What age range are you?" ✅ "What region do you live in?" ✅ "What's your household income bracket?"

❌ "What's your first and last name?" ❌ "What's your home address?" ❌ "What school do your children attend?" ❌ "Who is your employer?" ❌ "What is your phone number?"

Sensitive experiences and trauma disclosure

We allow research questions about sensitive personal experiences — including mental health, addiction, drug use, and medication. What we block is pressure for specific trauma disclosure: questions demanding details of specific abuse incidents, assault, or traumatic events.

✅ "Have you ever taken recreational drugs?" ✅ "Have you experienced addiction or substance dependency?" ✅ "Are you or have you ever taken anti-depressants?" ✅ "How has your mental health been over the past year?"

❌ "Describe in detail what your abuser did to you" ❌ "Tell me exactly what happened when you were assaulted"

OriginalVoices meta-questions

Questions about product methodology, sample sizes, or how the platform works.

❌ "How does OriginalVoices work?" ❌ "How can we trust the data without knowing the sample size?"

Nonsensical content

Incoherent or unreasonable questions that don't constitute valid research.

❌ "Purple elephant taxation banana?"

Non-English content

Questions must be in English. Other languages are not currently supported.

❌ "你喜欢这双破鞋的那喜芳面？" ❌ "Avez-vous envisagé des médicaments amaigrissants comme Ozempic?"

If your question is rejected

You'll be shown a notification explaining that the topic isn't permitted. You're welcome to rephrase your question to focus on:

• General experiences and opinions
• Behaviours and attitudes
• Non-sensitive topics or themes

Why these rules exist

These rules help us:

• Protect real people and their data
• Prevent harm or manipulation
• Comply with international safety and privacy standards
• Maintain trust and responsible use of Digital Twins
• Ensure insights remain ethical, respectful, and useful

Have feedback or need help?

We're still early, and these guidelines will evolve as we learn and grow. We'll continue reviewing and refining them to make sure they're fair, responsible, and appropriate for everyone using the platform.

If you're unsure whether a question is allowed, need help wording it, or have feedback on this policy, we'd love to hear from you: support@originalvoices.ai