CONTENT & COMMUNITY GUIDELINES

Wayward Fowl exists to create an independent public record of incidents, concerns, and harms involving artificial intelligence.

We want people to be able to report genuine experiences—including difficult, controversial, and disputed experiences—while maintaining a platform that is useful, fair, credible, and responsible.

These Content & Community Guidelines explain what we expect from people who use Wayward Fowl.

They supplement our Terms of Use, Incident Submission Terms, Privacy Policy, Legal Disclaimer, and Corrections and Dispute Policy.

1Our Core Principle

Report what happened.

You do not need to prove that an AI system caused an incident before reporting it.

You may report an experience if you reasonably believe artificial intelligence, an algorithmic system, automated decision-making, or related technology may have been involved.

But please distinguish among:

What you know

Facts you personally observed or can document.

What you were told

Information provided by another person or organization.

What you believe

Your interpretation of what happened.

What you suspect

A possibility that you cannot currently establish.

That distinction makes reports considerably more useful to everyone.

2Good-Faith Reporting

Reports should be submitted in good faith.

When submitting information, you should have a reasonable basis for statements you present as facts.

You do not need certainty about every aspect of an incident.

For example, it is appropriate to say:

“I believe an automated hiring system may have contributed to my rejection because…” when that accurately describes what you know.

It may not be appropriate to state:

“Company X intentionally discriminated against me using AI.”

as an established fact if you do not have a reasonable factual basis for that conclusion.

3Be Specific

Useful reports explain what actually happened.

Where reasonably possible, include information such as:

• what happened;
• when it happened;
• where it happened;
• the company or organization involved;
• the AI system or product involved, if known;
• how you know or suspect AI was involved;
• what decision or output occurred;
• what harm or consequence resulted;
• what the company told you;
• what steps you took afterward; and
• what evidence supports your account.

You do not need to know the technical details of an AI system to submit a report.

4Evidence Is Encouraged, Not Always Required

Evidence can make a report more useful.

Depending upon the incident, relevant evidence might include:

• screenshots;
• emails;
• letters;
• AI outputs;
• chat transcripts;
• decision notices;
• application records;
• transaction records;
• photographs;
• correspondence;
• policies;
• public records; or
• other documentation.

Not every incident will have documentary evidence.

The absence of documentation does not automatically prevent you from reporting an experience.

However, do not fabricate, materially alter, or deceptively present evidence.

5Protect Sensitive Information

Before uploading anything, review it carefully.

Please remove information that is not necessary to understand the incident.

In particular, do not unnecessarily publish:

• Social Security numbers;
• passwords;
• authentication codes;
• bank-account numbers;
• credit-card numbers;
• driver's-license numbers;
• passport numbers;
• private home addresses;
• personal telephone numbers;
• detailed medical records;
• biometric identifiers; or
• information identifying minors.

Wayward Fowl may redact sensitive information even if you do not.

6Report Systems And Organizations — Not Unnecessarily Individual Employees

Wayward Fowl is primarily interested in understanding AI systems, institutional practices, automated decisions, and their consequences.

Where possible, focus your report on:

• the AI system;
• company;
• organization;
• product;
• process;
• policy; or
• decision

rather than attacking individual employees.

You may identify an individual when their identity is genuinely relevant to understanding an incident, but avoid unnecessarily naming lower-level employees or other individuals.

7Harassment Is Not Permitted

Wayward Fowl may not be used to harass, threaten, intimidate, stalk, or target another person.

This includes:

• threats of violence;
• repeated targeted abuse;
• encouraging others to harass someone;
• publishing private information for purposes of intimidation;
• sexual harassment;
• threatening someone's family;
• extortion; or
• other unlawful harassment.

Criticism is permitted.

Strong criticism is permitted.

Reporting alleged misconduct is permitted.

Disagreement is permitted.

Harassment is not.

8No Knowingly False Reports

Do not knowingly submit information that is:

• fabricated;
• materially false;
• deliberately deceptive;
• impersonated;
• fraudulently manufactured; or
• intentionally presented in a materially misleading manner.

Making a mistake is not the same as intentionally submitting false information.

If you later discover a material error in your report, tell us so that it can be reviewed and, where appropriate, corrected.

9Facts And Opinions

Both factual reports and opinions may be valuable.

However, readers should be able to understand which is which.

Statements such as:

“The system rejected my application.”

are different from:

“I think the system was unfair.”

and different again from:

“The company deliberately designed the system to discriminate.”

The stronger the factual accusation, the more important it is to have a reasonable factual basis for making it.

10Discrimination And Fairness Reports

Wayward Fowl specifically welcomes good-faith reports concerning potential AI discrimination, bias, or disparate treatment.

These may concern characteristics or circumstances such as:

• race or ethnicity;
• sex or gender;
• age;
• disability;
• religion;
• national origin;
• language;
• socioeconomic circumstances;
• geography; or
• other characteristics relevant to an alleged incident.

Discussion of such characteristics may be necessary to understand potential AI harms.

However, the Platform may not be used to attack, demean, or promote hatred toward people based upon protected or personal characteristics.

11Types Of Incidents We Welcome

Wayward Fowl welcomes reports involving areas such as:

• employment and hiring;
• credit and lending;
• insurance;
• housing;
• healthcare;
• education;
• government services;
• criminal justice;
• surveillance;
• privacy;
• biometric systems;
• consumer products;
• customer service;
• autonomous systems;
• content moderation;
• generative AI;
• misinformation;
• impersonation and deepfakes;
• fraud;
• cybersecurity;
• accessibility;
• pricing;
• recommendations;
• algorithmic ranking;
• automated decision-making;
• AI agents;
• physical safety;
• reputational harm;
• economic harm; and
• other real-world consequences involving AI.

This list is not exhaustive.

12You Don'T Need To Know Which Ai System Was Used

People affected by automated decisions often are not told which system was involved.

That's okay.

If you know the system, identify it.

If you suspect a system but aren't certain, say so.

If you only know that the organization indicated that automation or AI was involved, describe what you were told.

If you don't know whether AI was involved but have a reasonable basis for suspecting it, explain why.

Wayward Fowl may later connect reports to particular systems as additional information becomes available.

13Confidential Information And Trade Secrets

Do not knowingly upload trade secrets or confidential information that you are legally prohibited from disclosing.

This may be particularly relevant if you are:

• an employee;
• former employee;
• contractor;
• vendor;
• researcher;
• consultant; or
• business partner

of the organization involved.

Wayward Fowl cannot determine for you whether a confidentiality agreement or other legal obligation permits a particular disclosure.

If you are uncertain, consider obtaining legal advice before submitting the material.

14Whistleblowers And Insiders

Wayward Fowl welcomes information from people with direct knowledge of AI systems and their effects.

However, submitting information to Wayward Fowl does not automatically give you legal whistleblower protection.

If you are considering disclosing confidential internal information or believe you may face retaliation, you may wish to consult an attorney before making a Submission.

Do not assume that Wayward Fowl can guarantee your anonymity.

15Information About Minors

Take particular care with information involving children.

Do not unnecessarily identify a minor publicly.

Wayward Fowl may redact names, photographs, locations, school information, or other identifying details concerning minors.

Protecting the child may take precedence over preserving every detail of a Submission.

16Sexual Or Graphic Content

Some legitimate AI-harm incidents may involve sexual content, deepfakes, exploitation, violence, or other disturbing material.

Such incidents may be reported.

However, only include graphic material when genuinely necessary to document the incident.

Wayward Fowl may:

• restrict access;
• add warnings;
• blur images;
• redact material;
• describe rather than display content; or
• decline to publish particular material.

Illegal sexual content involving minors is prohibited.

17Copyrighted Material

You may sometimes need to provide copyrighted material to document an incident.

Only submit material you have the right to submit or that you reasonably believe may lawfully be provided for the relevant purpose.

Wayward Fowl may remove material in response to valid intellectual-property complaints.

18Company And Organization Responses

Companies and organizations identified in reports are encouraged to participate constructively.

They may:

• dispute allegations;
• explain how a system works;
• identify factual errors;
• provide documentation;
• explain relevant policies;
• describe corrective actions;
• identify human involvement in a decision;
• explain why they believe AI was not involved; or
• provide other relevant context.

Responses are subject to these Guidelines.

A company response does not automatically invalidate the original report.

Likewise, publication of the original report does not mean Wayward Fowl has determined that the company's response is incorrect.

19Don'T Attack Reporters

Organizations and individuals may strongly dispute a report.

They may not use Wayward Fowl to:

• threaten the reporter;
• harass the reporter;
• publish unnecessary private information about the reporter;
• encourage retaliation;
• intimidate witnesses; or
• otherwise abuse the Platform.

Legitimate factual rebuttal is encouraged.

Retaliation through the Platform is not.

No Coercion or Manipulation of Reporters

Wayward Fowl does not permit individuals or organizations to use the Platform to improperly interfere with another person's decision to report an incident.

A company, employer, AI provider, organization, manager, or other person may not use Wayward Fowl to:

• require or pressure someone to submit an incident report;
• require or pressure someone to withdraw a report;
• require or pressure someone to materially alter a truthful report;
• require a person to disclose whether they submitted a report;
• require a person to reveal an anonymous or pseudonymous Wayward Fowl identity;
• demand proof that a person did or did not submit a report;
• offer compensation or another improper incentive in exchange for submitting a false or

misleading report;

• offer compensation or another improper incentive conditioned upon withdrawing or

materially changing a truthful report;

• threaten someone because they submitted or may submit a report; or
• organize or direct a campaign intended to artificially manipulate incident counts,

responses, ratings, classifications, or other Platform information.

This provision does not prohibit an organization from:

• asking Wayward Fowl to correct inaccurate information;
• submitting a response;
• communicating legitimate factual information;
• resolving an underlying dispute with a reporter;
• requesting that a reporter voluntarily correct information the organization reasonably

believes is inaccurate; or

• exercising legal rights through appropriate lawful processes.

Wayward Fowl may take action against accounts or organizations that attempt to manipulate the integrity of the incident-reporting process

20Reporters May Be Challenged

Submitting a report does not mean your account is immune from scrutiny.

Wayward Fowl may:

• ask questions;
• request evidence;
• contact relevant organizations;
• seek clarification;
• identify inconsistencies;
• publish responses;
• mark information as disputed;
• correct errors; or
• remove material where appropriate.

A credible public incident database requires meaningful scrutiny of information from all sides.

21Corrections Are A Good Thing

If you made a mistake, tell us.

Correcting an error does not necessarily undermine the rest of your report.

Wayward Fowl would rather maintain an accurate evolving record than pretend that every initial report is perfect.

Material corrections may be reflected in a correction history.

22Disputed Reports

A report may remain published even when a company or individual disputes it.

Where appropriate, Wayward Fowl may mark the report:

DISPUTED

and provide readers access to the competing account or additional information.

A disputed label does not mean Wayward Fowl has determined that either party is correct.

23Duplicate And Related Incidents

Multiple people may report similar experiences involving the same AI system or organization.

Wayward Fowl may:

• link related incidents;
• identify potential patterns;
• combine substantially duplicate submissions;
• create system-level or company-level pages; or
• aggregate incident data.

We will seek to avoid presenting multiple reports of the same underlying event as independent incidents where we know they are duplicates.

24Coordinated Manipulation

Do not attempt to manipulate Wayward Fowl by:

• submitting large numbers of fabricated reports;
• organizing false reporting campaigns;
• creating fake identities;
• artificially inflating incident counts;
• impersonating reporters;
• impersonating companies;
• submitting fraudulent evidence; or
• otherwise attempting to distort the public record.

Wayward Fowl may investigate suspected coordinated manipulation and remove associated content or accounts.

25Automated Submissions

Automated or bulk submissions are not permitted unless Wayward Fowl expressly authorizes them.

Researchers, advocacy organizations, companies, regulators, or others seeking to submit structured datasets should contact Wayward Fowl regarding appropriate data-submission methods.

26Commercial Promotion And Spam

Do not use incident reports primarily to:

• advertise products;
• promote services;
• generate leads;
• post affiliate links;
• manipulate search rankings;
• promote unrelated causes; or
• distribute spam.

Relevant references to products and services are permitted when genuinely connected to an incident.

27Political And Advocacy Content

AI incidents may involve government policy, elections, public officials, political organizations, or controversial public issues.

Good-faith reports concerning such incidents are permitted.

Wayward Fowl is an incident-reporting platform, not a general political discussion forum.

Content should have a meaningful connection to AI systems, automated decision-making, or related technology.

28Moderation

Wayward Fowl may moderate content to enforce these Guidelines and other Platform policies.

Depending upon the circumstances, we may:

• approve content;
• decline publication;
• request changes;
• redact information;
• add context;
• add labels;
• restrict visibility;
• remove content;
• suspend accounts;
• terminate accounts; or
• take other appropriate action.

Moderation does not necessarily mean that Wayward Fowl has verified the truth of the underlying report.

29Moderation Should Not Favor Companies Or Reporters

Wayward Fowl's goal is to maintain an independent record.

We should not remove a legitimate report merely because a powerful company dislikes it.

We also should not preserve information we determine should be corrected or removed merely because it was submitted by a reporter alleging harm.

Our policies apply to reporters, companies, organizations, and other participants.

30Commercial Relationships

Advertising, sponsorship, data licensing, insurance relationships, consulting relationships, or other commercial relationships should not determine how individual incident reports are moderated.

Where appropriate, Wayward Fowl may disclose material conflicts of interest.

31Enforcement

Not every violation will receive the same response.

We may consider factors such as:

• severity;
• intent;
• potential harm;
• history of violations;
• whether the problem can be corrected;
• public-interest value;
• privacy implications;
• safety concerns; and
• applicable law.

Possible actions include warnings, editing, redaction, removal, account restrictions, suspension, or termination.

32Appeals

Where Wayward Fowl provides an appeal mechanism, users may request reconsideration of certain moderation decisions.

An appeal should explain why the decision was incorrect and provide any relevant additional information.

Repeated appeals containing no material new information may be closed without further review.

33Reporting Content To Wayward Fowl

If you believe content violates these Guidelines, you may report it through:

info@waywardfowl.com If your concern involves factual accuracy or a report about you or your organization, use the process described in our Corrections and Dispute Policy.

34Legal Requests

Formal legal notices should be directed to:

Wayward Fowl Inc.

Attn: Legal

info@waywardfowl.com

35These Guidelines Will Evolve

Artificial intelligence changes quickly, and the types of incidents reported to Wayward Fowl will evolve.

We may update these Guidelines as the Platform develops.

When appropriate, we will explain material changes.

36The Standard We Ask Users To Follow

Before submitting something, ask yourself:

Is it relevant?

Do I have a reasonable basis for saying it?

Have I distinguished what I know from what I believe or suspect?

Have I removed unnecessary private information?

Would the report help someone understand what actually happened?

If so, it probably belongs on Wayward Fowl.

Reporting Principles

Be truthful.

Don't knowingly submit false information.

Be specific.

Tell people what happened rather than simply labeling someone or something as “bad.”

Show your work.

Provide evidence where you reasonably can.

Distinguish facts from conclusions.

Tell readers what you know and what you believe.

Protect privacy.

Don't expose information that isn't necessary.

Focus on systems and outcomes.

The purpose is to understand AI harms, not attack individuals.

Allow responses.

Companies and other affected parties should have a meaningful opportunity to present their side.

Correct mistakes.

An accurate evolving record is more valuable than an unchangeable one.

Disagreement is allowed.

Harassment, threats, fabrication, and manipulation are not.

Wayward Fowl Inc.

Community & Moderation: info@waywardfowl.com

Website: waywardfowl.com