Inspra Acceptable Use Policy — Version 1.2 — effective 12 October 2026

Inspra Acceptable Use Policy

Genius365 Pty Ltd, trading as Inspra — Version 1.2 — effective 12 October 2026

Key points

This policy is part of our Customer Terms. It applies to everyone using Inspra AI, including partners, their clients, and API and MCP users.
AI agents must say they are AI at the start of every conversation, answer truthfully if asked, and never claim to be human.
Some uses are never allowed, such as child exploitation, scams, impersonation, voice cloning without consent, political and electoral calls, tobacco and vape marketing and high-risk decisions made by AI without meaningful human review.
Some uses need our written approval first, such as financial products, credit, insurance, energy, debt collection and health.
Call and message lawfully, protect vulnerable people and never pressure anyone.
We act proportionately, explain why, restore service once things are fixed, and review decisions on request.
Report abuse, including unwanted calls from a number on our platform, to report-dnc@inspra.ai.

1. About this policy

a.
This Acceptable Use Policy (AUP) is published by Genius365 Pty Ltd ABN 50 679 963 967, trading as Inspra (we, us, our). It forms part of our Customer Terms (clause 4A). A breach of this AUP is a breach of clause 9(o).
b.
This AUP applies to you, to anyone using the platform through your account or access credentials (your users), to partners and their end clients, and to every sub-account and workspace. It covers use through the platform, our API, our MCP server and any connected tool or AI assistant. It also covers indirect use and attempts. You must not help or knowingly allow anyone else to do what this AUP prohibits.
c.
You are responsible for your users' compliance. Section 13 covers partners.
d.
This AUP adds to the Customer Terms, the Campaign and Batch Calling Terms (Campaign Terms) and the law. You must comply with all of them. If this AUP is inconsistent with the Customer Terms, the order of priority in clause 4(d) decides which prevails. A stricter rule of conduct in this AUP is not an inconsistency: it applies alongside the Customer Terms.
e.
Words defined in the Customer Terms have the same meaning here. In addition:
1. i.
  AI agent means a bot, or any other AI application on the platform that speaks or writes to people;
2. ii.
  conversation means a call, or an exchange of messages, between an AI agent and a target;
3. iii.
  partner means a customer with a partner agreement, and end client means a business that a partner supplies with our products; and
4. iv.
  section means a section of this AUP, and clause means a clause of the Customer Terms.
f.
This AUP still applies, and you remain responsible, when we set up an AI agent or supply a managed service for you (clauses 10.2 and 10.3). We follow this AUP in that work and in our own AI agents.

2. Core principles

These principles guide how this AUP is read. They are not separate rules: a breach must be of a specific rule in this AUP, the Customer Terms or the law.

a.
Lawful: use the platform lawfully.
b.
Honest: be open about AI, about who is calling, and why.
c.
Fair: respect people, take care with the vulnerable, and never pressure or exploit anyone.
d.
Respectful of privacy: use personal information only as the law and people's reasonable expectations allow.
e.
Safe: do not put people, networks or the platform at risk.
f.
Accountable: a person in your business must answer for each AI agent and be able to step in.

3. AI transparency

a.
Each AI agent must say that it is an AI at the start of every conversation. For example: 'Hi, this is Sam, an AI assistant calling for Smith Plumbing.' On inbound calls this goes in the greeting. On messaging channels it goes in the first message.
b.
If asked whether it is a person or an AI, the AI agent must answer truthfully. It must never claim or imply that it is human, for example by denying that it is an AI or by posing as a named member of staff. It may have a name, voice and personality.
c.
The AI agent must identify the business it acts for and, on outbound calls, the purpose of the call.
d.
Voicemail and other recorded messages in a synthetic or cloned voice must say that they are AI-generated and must name the business.
e.
This section applies even where no law requires the disclosure. It adds to clause 2.8 of Schedule 2 and to the Campaign Terms.

4. Voices, likeness and impersonation

a.
You may use a voice cloned from a real person only with that person's documented, informed consent (or, if they have died or cannot consent, the consent of someone legally entitled to give it). The consent must cover how, by whom and for how long the voice is used, and must allow withdrawal. Stop using the voice as soon as consent ends. Keep the consent record for at least 2 years after you stop using the voice, and give us a copy within 5 business days if we ask.
b.
A cloned voice is still an AI, so section 3 applies. The voice provider's terms also apply (clause 8A).
c.
You must not impersonate, or create audio, text or images that falsely appear to come from, a real person, a business you do not represent, a government agency or official (such as the Australian Taxation Office, Services Australia, the police or a court), a regulator (such as ACMA, ASIC, the ACCC or the OAIC), or a bank, telco or other provider the target deals with.
d.
You may run an AI agent for another business only with its written authority, and the AI agent must name that business truthfully.

5. Prohibited content and activities

You must not use the platform, an AI agent or any output for any of the following. We cannot approve any of these uses under section 6.

a.
Child safety: anything that sexualises, grooms, exploits or abuses a child, including child sexual abuse material. This ban is absolute. We will report suspected child sexual abuse material to the Australian Centre to Counter Child Exploitation or the police.
b.
Illegal goods and services: illegal drugs, the unlawful supply of weapons or explosives, counterfeit goods, human trafficking, regulated services you are not licensed to provide, or dealings in breach of Australian sanctions laws.
c.
Scams and fraud: scams, fraud, phishing, or any other deception to obtain money, goods or information.
d.
Social engineering: tricking or pressuring anyone into revealing passwords, one-time codes, card or bank details, tax file numbers, Medicare numbers or other government identifiers, or health information, or into giving remote access to a device.
e.
Misleading conduct: false or misleading statements about you, your goods or services, prices, a person's rights (such as cooling-off rights), urgency or who is calling, or fake reviews.
f.
Hate: promoting hatred, vilification or discrimination based on race, religion, sex, sexual orientation, gender identity, disability, age or a similar attribute.
g.
Harassment: threats, intimidation, stalking or abuse, or contacting people who have asked you to stop.
h.
Violent extremism: promoting terrorism or violent extremism, or inciting violence.
i.
Self-harm: promoting suicide, self-harm or eating disorders (see section 9(e)).
j.
Sexual content: sexually explicit content, adult chat or companion services, or arranging sexual services.
k.
Harmful misinformation: false information likely to cause harm, including about health, voting or emergencies.
l.
IP infringement: infringing IP or misusing confidential information, including in knowledge bases or by crawling websites you are not authorised to use.
m.
Privacy breaches: handling personal information unlawfully, including trading in personal information or leads without the consent or authority the law requires.
n.
Covert surveillance: recording, monitoring or tracking anyone without the notice or consent the law requires, or intercepting communications.
o.
Security attacks: malware, denial-of-service attacks, unauthorised access, or probing or attacking any system or network.
p.
Spam: commercial electronic messages (such as SMS, email and WhatsApp messages) sent without the consent, sender identification and unsubscribe facility that the Spam Act requires. Recorded and synthetic voice calls are covered by the Do Not Call Register Act and the Telemarketing Standard (section 10). Ringless voicemail needs our written approval (Campaign Terms clause B10(c)).
q.
Political and electoral communications: calls or messages for a political party, candidate, election, referendum or electoral campaign.
r.
Tobacco and vaping: advertising, promoting or selling tobacco, e-cigarettes or vaping products.
s.
Other unlawful conduct: anything else that breaches the law, or is likely to cause us to breach it (clauses 9(g) and 9(i)).

6. Restricted uses that need our prior written approval

a.
You need our prior written approval before using the platform for:
1. i.
  regulated goods: alcohol, therapeutic goods, gambling or wagering, or firearms and other weapons;
2. ii.
  financial products, credit or consumer leases;
3. iii.
  insurance;
4. iv.
  selling energy retail contracts, or obtaining consent to change energy retailer;
5. v.
  debt collection;
6. vi.
  health services in which an AI agent handles health information;
7. vii.
  charitable fundraising.
b.
To apply, email hello@inspra.ai describing the use, who you will contact, your scripts, your licences and your compliance controls. We will respond within 10 business days and decide reasonably. We may set reasonable conditions (such as script review, test calls or volume limits), which you must meet.
c.
We will not approve gambling marketing to people registered on the National Self-Exclusion Register (BetStop), any use that needs a licence you do not hold, or any use that section 5 prohibits.
d.
An approval covers only the use described. It is not legal advice or confirmation that the use is lawful.
e.
If we approve a use involving health information, your account is opted out by default of our use of de-identified call data (clause 11.2(f)).
f.
We may vary or withdraw an approval on 30 days' notice with reasons, or sooner if needed to comply with the law or a carrier or regulator direction, or to prevent serious harm. If we withdraw an approval when you are not in breach, and the approved use is essential to a product you have paid for, you may end that product and clause 6.4(b) of the Customer Terms applies to your refund.
g.
If you are already carrying on a restricted use when this AUP first applies to you, you may continue it for 60 days. If you apply within that time, you may continue until we decide your application.

7. High-risk AI uses we do not allow

You must not use the platform for:

a.
Automated decisions: decisions with a legal or similarly significant effect on a person, unless a person with the authority and information to change the decision meaningfully reviews it before it takes effect. This covers decisions about credit, employment, insurance, housing, essential services, government benefits and health care (including triage and diagnosis). An AI agent may still collect information, answer questions and book appointments.
b.
Emergency and life-safety uses: emergency services, crisis or suicide-prevention lines, medical or personal alarms, or any means of reaching emergency services. An AI agent cannot reliably connect anyone to 000 (clauses 8.6 and 9(j)).
c.
Biometrics: identifying or verifying people by voice or other biometrics, or creating voiceprints, without their express consent. Inferring race, religion, sexuality, health or other sensitive characteristics from a person's voice or features is never allowed.
d.
Emotion inference: using emotion, stress or 'lie detection' analysis to make decisions about a person or to aim persuasion at their emotional state. Sentiment analysis for quality assurance, coaching or detecting distress is allowed.
e.
Targeting vulnerability: targeting people because of age, illness, disability, hardship, addiction, grief or limited English in order to sell to them or gain from them. Services that help such people and meet section 9 are allowed.
f.
You must give any notices about automated decision-making that the law requires of you, including in your privacy policy from 10 December 2026 (APP 1.7 to 1.9).

8. Professional advice

a.
An AI agent must not give medical, legal or financial advice (including financial product advice, credit assistance or tax advice) unless three conditions are met: you hold the licence or registration the law requires, the advice is within its scope, and we have approved the use under section 6. Even then, the AI agent must offer a qualified person where appropriate.
b.
Otherwise, an AI agent may give factual information from your approved materials, such as opening hours, prices, product features or booking details. It must not make recommendations about a person's circumstances.

9. Vulnerable people

Any target may be vulnerable, for example because of age, illness, disability, mental health, family violence, bereavement, financial hardship or limited English. Your AI agents and campaigns must:

a.
recognise signs of vulnerability, such as confusion, distress, or mention of illness, hardship or a carer;
b.
when those signs appear, offer a person or a call back, or end the conversation politely, and stop any sales or collection attempt;
c.
never pressure anyone, such as through false urgency, pushing after a person says no, or seeking agreement from someone who does not appear to understand;
d.
when a target mentions financial hardship, stop sales or collection activity and give your hardship contact details or transfer them to a person (you must also meet any hardship obligations that apply to you);
e.
if a target is in danger or at risk of self-harm, tell them to call 000 in an emergency, give crisis support details such as Lifeline (13 11 14), and hand over to a person where one is available;
f.
end any sales, marketing or collection activity on learning that a target is under 18; and
g.
let targets ask for a person, with a practical way to provide one or to call back.

Keep records of how vulnerability was handled, and use them to improve your campaigns.

10. Calling, messaging and telephony

This is a summary. The Campaign Terms set out the detail, and clause 8 covers numbers. You are responsible for how numbers are used, caller ID and calling compliance (clauses 8.3 and 8.5). You must follow these rules:

a.
Consent: have a lawful basis, and any consent the law requires, for every call and message. Check lists against the Do Not Call Register and your suppression list as the Campaign Terms require, and keep evidence of consent.
b.
Hours: call only within the hours allowed by the Telecommunications (Telemarketing and Research Calls) Industry Standard 2017, in the target's local time.
c.
Caller ID: show a valid caller ID that you are authorised to use, that identifies you, and that can take return calls for at least 30 days. Never spoof or hide it.
d.
No number cycling or generation: never rotate numbers to avoid blocking, spam labels or complaint detection, and never dial numbers generated in sequence or at random. You may use separate numbers for genuinely separate brands, business units or regions.
e.
Disclosures: at the start of each call, identify the business and the purpose of the call, say that the target is speaking with an AI, and disclose any recording or transcription. Recording settings are fixed before an AI agent or campaign goes live, and recording cannot be stopped during a call, so if a person objects to recording you must end the call.
f.
Opt-outs: end a call when the person asks, honour opt-outs across all your campaigns, numbers and sub-accounts, and action unsubscribe requests for messages within 5 business days.
g.
Messaging: use only sender IDs you are entitled to use, and make sure any alphanumeric sender ID is registered on the ACMA SMS Sender ID Register. Follow Meta's WhatsApp Business policies.
h.
Restricted numbers: never let an AI agent call 000, 112, 106 or another emergency number. Do not call premium-rate, satellite or high-cost international numbers unless we have enabled them.
i.
No traffic pumping: never inflate traffic, make calls to earn termination or revenue-share payments (including international revenue share fraud), or allow toll fraud. Tell us at once if you suspect any of these.
j.
List hygiene: remove disconnected, invalid and opted-out numbers, and keep failed and very short calls within the thresholds in the Campaign Terms. High failure rates burden carrier networks and can get your traffic blocked.
k.
Carrier rules: follow the carrier rules and directions we pass on to you under clause 8.7.

11. Platform integrity and security

a.
You must not, and must not help anyone to:
1. i.
  reverse engineer the platform, or extract its code, models, weights, system prompts, templates or configurations, other than your content (clause 9(d));
2. ii.
  scrape the platform, or access it by automated means other than our documented API and MCP server;
3. iii.
  jailbreak, prompt-inject or otherwise bypass safety features, guardrails, usage or rate limits, verification or billing controls;
4. iv.
  share one login between several people, or let anyone other than your users use your account (clause 9(c)). Each user needs their own login where the platform supports it;
5. v.
  use extra or new accounts to avoid verification, limits, charges or enforcement (clause 9(n));
6. vi.
  load test, stress test or penetration test the platform or a connected carrier without our written permission;
7. vii.
  publish benchmarks of the platform without our consent, which we will not unreasonably withhold (honest reviews are always allowed);
8. viii.
  use the platform, its outputs or call data to build or train a competing product or model (clause 9(a)); or
9. ix.
  interfere with other customers or their data.
b.
You must keep your account secure, give users only the access they need, and tell us at once if credentials may have been compromised. You are responsible for use of your access credentials by anyone you allow to use them, including connected tools and AI assistants, and by anyone who gains access because you did not keep them secure, except to the extent our breach or negligence caused it (Customer Terms Schedule 1 clause 3(l)).
c.
Responsible disclosure: report security vulnerabilities to compliance-officer@inspra.ai. Do not access other people's data or disrupt the service, and give us reasonable time to fix the issue before telling anyone else. Good-faith research that follows these rules is not a breach of this AUP.

12. Fair use and capacity

a.
Your plan or order sets your platform minutes, AI agents and concurrent calls. Your carrier supplies and charges for phone numbers separately. Concurrency limits differ by plan. No plan is unlimited.
b.
You must not overload the platform or our carriers, for example with a sudden, very large burst of calls. Tell us before you run a campaign much larger than usual.
c.
We may temporarily throttle call, concurrency or message rates where reasonably needed to protect the platform, carriers or other customers, to manage capacity, or to meet a carrier or regulator requirement. We will give notice where practicable, use the smallest limit that works, and lift it when the reason has passed. Throttling is not by itself a finding of breach. If we throttle your account for more than 2 business days for a reason you did not cause, we will not charge plan fees for that period and will extend the expiry date of your paid credit by the same period.
d.
If your usage changes suddenly in a way that suggests your account has been compromised, we may pause that activity and contact you straight away.

13. White-label partners and end clients

a.
A partner is responsible to us for its end clients' use as if it were its own, within the liability terms of the Customer Terms and its partner agreement. Section 3 applies to every end client, whatever the branding.
b.
Each end client must be a business carrying on business in Australia with a valid ABN. Before an end client goes live, you must verify its identity, its business, its ABN and its intended use. Give us these records within 5 business days if we ask.
c.
Your agreement with each end client must include terms at least as protective as this AUP (by reference or under your own brand). It must also let you suspend or end the end client's access, allow monitoring under section 14, require cooperation with regulators and carriers, and promise nothing this AUP does not allow.
d.
You must enforce those terms promptly. If we ask you to act on an end client, you must do so within the time we reasonably set, which will be at least 1 business day unless section 15(c) applies. You must pass complaints, abuse reports and notices to the end client without delay.
e.
If you do not act, or section 15(c) applies, we may act directly on an end client's workspace, numbers or campaigns. We will act at the end-client level unless the problem affects your whole account. We will tell you first where practicable, and we may contact the end client directly.
f.
A partner agreement does not reduce the protections this AUP gives to the people that AI agents interact with (clause 5(a)).

14. Monitoring

a.
We may, but are not obliged to, monitor use of the platform. We do so to check compliance with this AUP, the Customer Terms, the Campaign Terms and the law, and to protect people, the platform and carriers.
b.
We may use automated tools (such as content classifiers, call-pattern and complaint analytics, and number-reputation data), including tools from providers on our Sub-processor List. Authorised staff may review campaign designs, lists, call data, recordings and transcripts, and make test calls.
c.
We monitor only as reasonably needed and limit access to authorised staff bound by confidentiality. We handle personal information as our Privacy Policy explains, and we never use monitoring to train AI models (clause 11.2(d)).
d.
Monitoring does not mean that we approve or have checked your campaigns. It does not shift your responsibility for compliance to us (clause 12.1 and Schedule 2).

15. Enforcement

a.
If we reasonably believe this AUP has been breached, we may take proportionate steps, starting with the least restrictive step that will work:
1. i.
  a warning;
2. ii.
  requiring changes to a campaign design, script, list, number or setting;
3. iii.
  throttling;
4. iv.
  pausing an AI agent, campaign, number or sender ID;
5. v.
  suspending an end client's workspace, a sub-account or your account (clause 15.3); or
6. vi.
  ending your licence (clause 15.2).
b.
Unless section 15(c) applies, we will tell you about the breach and give you a reasonable chance to fix it before we pause, suspend or end anything. Ending your licence is always subject to clause 15.2.
c.
We may act immediately only on the grounds in clause 15.3(b): where we reasonably believe it is needed to prevent serious harm to anyone, because the activity is or is likely to be unlawful, to comply with a direction or requirement of a carrier, regulator, law enforcement agency or court, or to protect the security or integrity of the platform or a carrier's network. We will take only the steps needed, and we will tell you promptly afterwards.
d.
We will tell you what we did, why, and how to have it lifted, unless the law or an authority prevents us. We will lift it promptly once the breach is fixed or found not to have happened. Clause 15.3 applies, including that we do not charge plan fees for a suspension that turns out to be unjustified.
e.
You must pay the reasonable costs we actually incur, to the extent they are caused by your breach. These include carrier charges or penalties passed on to us (clause 8.8) and reasonable third-party investigation costs. We will itemise these costs if you ask. We will reduce them to the extent we contributed, charge no penalty or fee for the breach itself, and not recover the same amount twice. A regulator's fine can be recovered only under clause 3 of Schedule 2.

16. Reporting abuse

a.
Anyone can report misuse to report-dnc@inspra.ai or on 1300 INSPRA (1300 467 772). This includes a member of the public who gets an unwanted or suspicious call or message from a number on our platform. Please include the number that contacted you, your number, the date and time, and what happened.
b.
We will acknowledge a report within 5 business days, pass it to the customer using the number (and its partner, if any) and act under section 15 where needed. We will share your phone number so the customer can stop contacting you, but not your name or other details without your consent unless the law requires it. We may not be able to tell you what action we took.
c.
If we pass a report to you, you must respond within 5 business days.
d.
You can also contact ACMA (telemarketing, the Do Not Call Register and spam), Scamwatch (scams), the OAIC (privacy) or the police (threats). In an emergency, call 000.

17. Cooperation with regulators, law enforcement and carriers

a.
We cooperate with law enforcement agencies, ACMA, the OAIC, the ACCC, ASIC, the eSafety Commissioner, other regulators, courts and carriers. This includes carrier traceback and scam blocking under the Reducing Scam Calls and Scam SMs Industry Code (C661).
b.
We may disclose information about your use where the law requires or permits it, or where it is reasonably necessary to prevent serious harm, as our Privacy Policy explains. We will tell you where the law allows.
c.
While a complaint, investigation or legal request is open, we may preserve relevant records, including by suspending deletion under your retention settings.
d.
You must cooperate with us, and with any regulator, agency or carrier investigating your use. You must provide consent records, lists, scripts and call records within 5 business days after we ask (or sooner if required), and you must not delete relevant records.

18. Asking us to review a decision

a.
You may ask us to review any step we take under this AUP, including refusing or withdrawing an approval. Email hello@inspra.ai with 'AUP review' in the subject line.
b.
We will give you our decision and reasons in writing within 5 business days (or sooner where the Campaign Terms set a shorter time for a paused campaign or number), or tell you within that time why we need longer and when we will respond. Where practicable, someone not involved in the original decision will do the review.
c.
The step may stay in place during the review only while the reason for it continues. If it was not justified, we will reverse it promptly, and clause 15.3 applies to plan fees.
d.
A review does not affect your rights under clause 16A (disputes) or at law. End clients should ask their partner, who may ask us.

19. Other terms

a.
No third-party rights: no one other than you and us, including a target, end client or user, may enforce this AUP. This does not limit anyone's legal rights.
b.
No waiver: if we do not act on a breach, that does not mean we accept it.
c.
Changes: we may change this AUP only under clause 14. A materially adverse change needs 30 days' notice by email (or as much notice as practicable if it is urgent), and you may then end your licence with a refund under clause 14(c). We publish each version with its effective date and keep earlier versions at https://inspra.ai/legal.
d.
Interpretation: clause 22.2 of the Customer Terms applies to this AUP.

20. Contact us

Genius365 Pty Ltd ABN 50 679 963 967, trading as Inspra, Level 10, 3 Bowen Crescent, Melbourne VIC 3004.

General enquiries, approvals and reviews: hello@inspra.ai or 1300 INSPRA (1300 467 772)
Abuse reports and unwanted calls: report-dnc@inspra.ai
Security and privacy: compliance-officer@inspra.ai
Legal notices: under clause 16 of the Customer Terms (email hello@inspra.ai with 'Important Legal Notice' in the subject line)
Our policies: https://inspra.ai/legal