AI Disclosure

Social Butterfly includes Butterfly AI, an assistant that drafts and rewrites social media content for you. This page explains how it works, what it is good at, where it falls short, and what we need you to do before you publish anything it produces.

1. What Butterfly AI is

Butterfly AI is a generative artificial intelligence feature built into Social Butterfly. Depending on your plan, it can:

• Draft captions, post copy and hooks from a short prompt.
• Rewrite something you have written in a different tone, length or format.
• Adapt one post for the conventions of several platforms.
• Suggest hashtags, headlines, content ideas and posting themes.
• Summarize comments or analytics, and suggest replies.

Using Butterfly AI is always your choice. Nothing is generated unless you ask for it, and nothing is published unless you schedule or post it.

2. How it works

Butterfly AI runs on large language models provided through Microsoft Azure OpenAI Service. When you make a request, we send your prompt plus the context needed to answer it — which may include your brand voice settings, the draft you are working on, and relevant recent post performance — to that service, which returns generated text.

These models work by predicting likely sequences of words based on patterns in the data they were trained on. They are not databases, they do not look things up, and they have no understanding of whether what they produce is true. That single fact explains most of the limitations described below.

Under Microsoft's terms for Azure OpenAI Service, your prompts and the generated responses are not used to train the underlying models and are not shared with other customers. We do not use your prompts or outputs to train any model of our own. Our Privacy Policy covers how we handle this data.

3. What Butterfly AI cannot be relied on for

Please treat every output as a first draft written by an enthusiastic assistant who has not checked their facts. In particular:

• It can be confidently wrong. AI models invent facts, statistics, dates, quotes, prices, product features, names and citations, and present them in a fluent and convincing way. This is a known and unavoidable property of the technology.
• It does not know what happened recently. The models are trained on data up to a cut-off date and have no live knowledge of current events, news, prices, product launches or platform rules.
• It does not know your business. Beyond what you tell it, it does not know your actual products, policies, prices, availability, terms or commitments. It will fill gaps with plausible-sounding guesses.
• It may reflect bias. Training data carries the biases of the material it was drawn from, and output can reflect stereotypes or leave groups out.
• It may resemble existing work. Output is not guaranteed to be original. It may closely match phrasing that already exists, including material protected by copyright or trademark.
• It is inconsistent. The same prompt can produce different answers each time. A good result once is not a guarantee of a good result next time.
• It does not know the rules that apply to you. It cannot reliably apply advertising law, disclosure requirements, industry regulations or a platform's community guidelines.
• It is not professional advice. Output is never legal, medical, financial, tax or other professional advice, whatever it may sound like.

4. No guarantee of accuracy or suitability

To the fullest extent permitted by law, Synergy Relations LLC makes no representation or warranty of any kind about content generated by Butterfly AI. We do not warrant that it is accurate, current, complete, reliable, original, non-infringing, unbiased, compliant with any law or platform policy, or fit for any particular purpose. AI output is provided "as is."

We do not review AI output before it reaches you, and we do not review or approve what you publish. Your use of AI-generated content is at your own risk, and the limitations of liability in our Terms of Service apply to it in full.

5. What we need you to do

Before publishing anything Butterfly AI produced, you agree to:

1. Read every word. Never schedule or publish AI output unread.
2. Check every factual claim — statistics, dates, prices, product details, quotations, names, awards, scientific and health claims — against a source you trust.
3. Confirm it is true of your business. Make sure any promise, offer, guarantee or capability described is one you actually stand behind.
4. Check it does not copy anyone. Search distinctive phrases, slogans and taglines before using them, and make sure you are not using someone else's trademark.
5. Apply the rules that apply to you. Include any required disclosures — sponsorship, affiliate and endorsement disclosures under FTC rules, and any industry-specific disclaimers for health, financial, legal, alcohol, cannabis, firearms, gambling or political content.
6. Follow platform rules on AI content. Several platforms require AI-generated or AI-modified content to be labeled. Those requirements change often, and meeting them is your responsibility.
7. Check tone and sensitivity. Make sure the content is appropriate for your audience, and re-read anything scheduled in advance if the news environment changes.
8. Make it yours. Edit output so it sounds like you rather than like a machine.

6. Things you must not use Butterfly AI for

In addition to the acceptable use rules in our Terms of Service, you may not use Butterfly AI to generate:

• Content that impersonates a real person, brand or organization, or that fabricates a quote or endorsement.
• Deceptive claims about a product, service, price, result or qualification.
• Fake reviews, testimonials, or content presented as coming from an independent customer.
• Deliberate misinformation about elections, public health, or emergencies.
• Harassing, defamatory, hateful or violent content.
• Sexual content involving minors, or any content that sexualizes or endangers children.
• Content that infringes copyright, trademark, or a person's right of publicity.
• Spam, mass unsolicited messaging, or content designed to manipulate platform algorithms.
• Anything intended to evade a platform's moderation, safety systems or account restrictions.

You also may not use Butterfly AI output to train, fine-tune or evaluate another machine learning model. We may suspend AI features, or your account, for violations.

7. Who owns AI-generated content

As between you and Synergy Relations LLC, you own the content Butterfly AI generates for you, to the extent ownership rights exist in it, and you are free to use it commercially.

Two honest caveats. First, because these models are statistical, we cannot promise your output is unique — a similar prompt from someone else may produce similar text. Second, U.S. copyright law generally protects only work with human authorship, so purely machine-generated material may not be protectable by copyright at all. Your own edits and creative choices are what make a piece defensible as yours. Neither point stops you using the content; both are worth knowing before you build a brand asset on top of it.

8. Human oversight and automated decisions

Butterfly AI assists with drafting. It does not make decisions about people, and we do not use it to make or support decisions that produce legal or similarly significant effects on anyone — nothing in the service decides on employment, credit, housing, insurance, education or healthcare. Every AI output passes through a human being — you — before it can reach an audience. There is no setting that publishes AI content without a person approving it.

9. Availability and changes

AI features depend on our provider's service and on usage limits that may apply to your plan. They may be unavailable, slow or rate-limited at times, and we may change, restrict or discontinue them, or change which models sit behind them. Model changes can alter the style and quality of output, so please re-check results after an update rather than assuming past behavior continues.

10. Reporting a problem

If Butterfly AI produces something harmful, biased, plainly false, or that appears to copy existing work, please tell us at [email protected]. Include the prompt and the output if you can. Reports genuinely help us improve safeguards.

11. Contact us

This AI Disclosure forms part of our Terms of Service. Using Butterfly AI means you accept it.