Research, Experiment & High-Risk Innovation Safety Policy

Plain-language summary

Big, bold, “impossible-today” ideas are welcome. We never reject a serious idea just because it cannot be built yet.
Instead of saying “impossible,” we ask: what breakthroughs would be needed? We label ideas by breakthrough level, safety risk, and research maturity.
The idea is welcome; dangerous step-by-step build instructions are not. We restrict content that could directly help someone cause real harm.
Some topics are prohibited or strictly reviewed: weapons, explosives, chemical or biological or radiological or nuclear hazards, malware and cyber-abuse, fraud, unsafe human experiments, and more.
Nothing here is professional engineering, medical, legal, or financial advice. Always follow the law, regulators, your institution, and qualified professionals.
High-risk ideas, projects, and events are escalated to human administrators for review.

1. Our mission: speculation is welcome

Clap Ideas is a future-technology lab. Our mission is Democratizing Ideas. Building the Future.Serious speculative ideas are welcome even if they seem impossible with today’s science. Visionary thinking is the point of this platform.

We never reject a genuine idea simply because it cannot be built right now. Instead, we reframe the question. Rather than saying “that is impossible,” we ask: what scientific or engineering breakthroughs would be needed to make it real? This turns a dead end into a research direction.

This policy protects that mission. It draws a clear line between imagining and discussing an idea (welcome) and giving usable instructions that could cause real-world harm (restricted).

2. How we classify ideas, not censor them

Every idea on the Platform can be classified along three independent dimensions. These labels describe an idea. They do not judge whether you are allowed to think about it.

Dimension	What it describes
Breakthrough level	How far beyond today’s technology the idea is, and what new discoveries it would need. From “buildable now” to “needs a major scientific breakthrough.”
Safety risk	How much real-world harm the idea could cause if someone tried to build it without proper controls. Most ideas are low risk; a few are high risk.
Research maturity	How much real evidence, prior work, or testing exists. From pure speculation to established science.

A high breakthrough level is not a problem. A high safety risk is what triggers extra review. An idea can be wildly ambitious and still perfectly welcome, as long as it does not include dangerous instructions and does not fall into a prohibited category.

3. The idea is welcome; the recipe may not be

This is the most important rule in this policy. There is a difference between:

Discussing a concept— “Could we one day capture energy from lightning?” This is welcome.
Giving a usable harmful recipe — precise steps, quantities, materials, or code that would let someone build a weapon, a dangerous device, a pathogen, or a cyber-attack. This is restricted, even if the surrounding idea is interesting.

You can explore almost any frontier in terms of goals, science, and trade-offs. You must not post operational instructions that could enable real harm. If your idea needs dangerous detail to be discussed, describe the goal and the open research questions instead of the dangerous method.

4. Prohibited and strictly reviewed categories

The following categories are prohibited or are strictly reviewed by human administrators. For these topics, you may describe a high-level goal or research question, but you must not provide instructions, designs, code, formulas, quantities, or procurement guidance that could enable harm.

4.1 Weapons, explosives, and CBRN hazards

We prohibit instructions for weapons, ammunition, explosives, or incendiary devices, and for chemical, biological, radiological, or nuclear (CBRN) hazards. This includes designs, synthesis routes, enrichment methods, dispersal methods, or anything that helps someone build, acquire, or use such hazards.

4.2 Malware, cyber-abuse, and surveillance abuse

We prohibit working malware, ransomware, exploit code intended to attack real systems, credential theft, account-takeover tooling, fraud kits, and instructions for unlawful hacking, denial-of-service attacks, or stalkerware. We also prohibit abusive surveillance of people without lawful authority and consent.

4.3 Fraud and financial crime

We prohibit schemes and instructions for fraud, scams, phishing, money laundering, payment fraud, counterfeit goods or documents, and other financial crimes.

4.4 Unsafe human-subject experimentation

We prohibit proposals or projects that experiment on people without lawful, independent ethical review and informed consent. Real human-subject research requires an ethics board or institutional review board, regulatory approval where applicable, and proper safeguards.

4.5 Medical diagnosis, treatment, and clinical activity

We prohibit medical diagnosis or treatment claims, prescriptions, dosing instructions, or clinical activity by people who are not authorised to provide them. Ideas about future medicine are welcome; practising medicine without authorisation is not.

4.6 Unsafe physical, aerospace, robotics, and energy experiments

We strictly review ideas that involve dangerous physical experiments, including high voltage, high pressure, high temperature, radiation, uncrewed aircraft in restricted airspace, autonomous robotics that could injure people, untested propulsion, and risky energy or transport experiments. Describe the science and safety questions, not a dangerous build procedure.

4.7 Dangerous step-by-step instructions of any kind

Beyond the categories above, we restrict any step-by-step instructions whose main use is to help someone cause serious physical, digital, financial, or environmental harm.

4.8 Extremism, violence, self-harm, and exploitation

We prohibit content that promotes terrorism, violent extremism, or violence against people; content that encourages self-harm or suicide; and any sexual content involving minors or the exploitation of any person. If you or someone else is in danger, see our Safety & Abuse Reporting Policy.

4.9 Illegal procurement, sanctions, and export-control evasion

We prohibit guidance on acquiring controlled or illegal items, evading sanctions, or breaking export-control laws.

5. Dual-use caution

Many technologies are dual-use: they have good uses and harmful uses. Energy, biology, chemistry, robotics, and software can all help or hurt people depending on how they are used. When an idea is dual-use, we ask you to focus on the beneficial purpose and the open research questions, and to leave out details that mainly help with the harmful use.

When in doubt, write less detail, not more. You can always describe the goal, the science, and what would need to be proven, without handing over a dangerous recipe.

6. No professional advice

Nothing on Clap Ideas is professional engineering, medical, legal, financial, scientific, or safety advice. AI-generated research briefs, scores, and suggestions may be inaccurate or incomplete. See our AI Transparency Policy for what AI can and cannot do.

Before acting on any idea in the real world, get advice from qualified professionals and follow the proper rules. Do not treat content here as a green light to build anything.

7. Your responsibilities

If you take any idea beyond discussion, you are responsible for following:

All applicable laws and regulations.
Directions from regulators and government authorities.
The rules of any institution, employer, school, or lab you work with.
Guidance from qualified professionals in the relevant field.
Recognised safety standards and codes of practice.
Ethics boards and independent review where people, animals, or the environment are involved.

You are responsible for what you build and do offline. The Platform hosts a discussion; it does not supervise, approve, or insure your real-world work.

You can still explore frontier topics responsibly. When an idea is sensitive:

Describe the goal and the open questions, not a usable harmful method.
Talk about what breakthroughs are needed rather than how to cause harm.
Leave out exact materials, quantities, code, formulas, and procurement steps for restricted topics.
Add a clear note about the risks and the safety or ethical review the idea would require.

9. AI review and human escalation

Clap Ideas follows an AI-autonomous, human-governed model. AI moderation pre-checks every submission, classifies safety risk, and recommends an action. AI can label an idea, lower its ranking, hide it, or queue it for review. For high-risk ideas, projects, and events, AI escalates to human administrators, who make the final decision.

Human administrators handle removals, account actions, appeals, and any dangerous-content or legal issue. See our Moderation, Enforcement & Appeals Policy for the full process and how to appeal.

10. High-risk projects and events

If an idea becomes a project or an event, extra rules apply. User-led projects and events are independent, and organisers are responsible for legality, safety, and participant welfare. High-risk activity needs proper authorisation and safeguards before it goes ahead. See our Projects, Teams & Events Terms.

11. Report unsafe content

If you see dangerous instructions, an unsafe project or event, or any prohibited content, please report it. Use the report tools described in our Safety & Abuse Reporting Policy. If anyone is in immediate danger, contact your local emergency services first.

For legal or sensitive safety questions, you can also contact legal@clapideas.com or security@clapideas.com.

This policy works together with our Acceptable Use Policy, Projects, Teams & Events Terms, Moderation, Enforcement & Appeals Policy, and AI Transparency Policy. For questions, contact legal@clapideas.com.

Change history is tracked by document version; see the Legal Centre.