STATUS. This is the Acceptable Use Policy referred to in clause 1.1 of the Omnistra Software as a Service Agreement (OMN-LGL-SAA-2026.1). It forms part of that Agreement and binds both Parties. Words defined in the Agreement have the same meaning here. This version is not amended; a correction is issued as a new version at a new permanent address, and clause 22 of the Agreement governs when a new version applies to a Customer.

1. Purpose and Standing

1.1The Acceptable Use Policy states the permitted and prohibited uses of the Services.
1.2It includes restrictions which a Supplier requires Omnistra to pass on.
1.3Clause 14.6 applies to the Acceptable Use Policy.
1.4Scope. The Customer’s obligations under this Policy relate to its own acts, omissions, Customer Content, configuration and instructions, and to Output resulting from them. Output falling within clause 10.6 is not a breach by the Customer, save that Omnistra’s rights under clause 14.4, under paragraph 4.1 of this Policy and under paragraph 3.1(p) are exercisable in respect of any Output within paragraph 2 or paragraph 3.1(i), (p), (q), (r) or (t) whether or not the Customer is in breach; and where Omnistra terminates on that footing the termination takes effect as a termination under clause 21.2 and clause 21.6 applies to it.

2. Prohibited Sectors and Subject Matter

2.1The Customer shall not use the Services, or configure the AI Agent, to sell, promote, advertise, display for sale, arrange, facilitate, advise on or discuss:
1. (a)Drugs, medical devices and health claims. Any drug, medicine or medical device requiring a licence, registration or approval the Customer does not hold; and any claim of therapeutic, curative, preventive, diagnostic or disease-modifying effect, or any other health or treatment claim, made without the prior approval of the Licensing Authority where that approval is required. Registered cosmetics, skincare, supplements and food products may be sold and described if no such claim is made and the Customer holds every licence required.
2. (b)Legal advice. Legal advice, legal opinions, or the drafting of legal documents. The AI Agent shall decline and refer.
3. (c)Financial services. Investment or securities advice; the offering, promotion or arranging of credit, loans, deposits, insurance or mobile financial services products; the quoting of returns. Transactional payment instructions for the Customer’s own goods are permitted if the Customer complies with the rules applicable to that payment method.
4. (d)Debt collection. The collection or recovery of debt on behalf of any person other than the Customer, on any Channel. The Customer may communicate with an End Customer about a sum owed to the Customer if the communication complies with law and the applicable Channel policy, is made between 09:00 and 20:00 Bangladesh Standard Time, is made not more than once in any 24-hour period and not more than three times in any seven-day period, ceases on the person’s request that it cease, and does not threaten legal or criminal process the Customer does not intend to take, disclose the debt to any other person, or use abusive, humiliating or intimidating language.
5. (e)Gambling and betting. Any gambling, betting or gaming for money, and the advertising, promotion, sponsorship, affiliate marketing, endorsement or financial facilitation of it.
6. (f)Political and electoral messaging. Political, electoral, referendum, party, candidate or advocacy messaging; campaigning; lobbying; and any use during an election period which promotes or opposes a candidate, a party or a referendum proposition. "Election period" means the period from announcement of the election schedule by the Election Commission to declaration of the result.
7. (g)Adult goods and content. Sexually explicit or sexually suggestive goods, content or persona configuration; and any romantic, companion or emotional-support persona.
8. (h)Weapons, controlled and illegal goods. Firearms, ammunition, explosives and weapons; tobacco and alcohol except where the Customer holds every licence required and the Channel permits it; endangered species and products derived from them; counterfeit, pirated, stolen, expired, unregistered or otherwise prohibited goods; and intoxicants.
9. (i)Multi-level marketing. Multi-level marketing, pyramid schemes, chain schemes and comparable business models.
10. (j)Live animals, body parts and unregulated healthcare items. Live animals; human or animal body parts, organs, blood, tissue and bodily fluids; and any healthcare, medical or wellness item the sale of which is not permitted by sub-paragraph (a). This sub-paragraph does not restrict the sale of registered cosmetics, skincare, supplements and food products permitted by sub-paragraph (a), and paragraph 3.1(o) applies in addition to it.

3. Prohibited Conduct

3.1The Customer shall not, and shall not permit any person to:
1. (a)remove, delay, obscure, contradict, or instruct the AI Agent to deny, the AI identity disclosure required by clause 10.3 or the recording announcement required by clause 11.2;
2. (b)configure the AI Agent to impersonate any identifiable natural person, business, government body, bank or mobile financial services provider, or to present itself as a named human employee;
3. (c)create or request a synthetic voice derived from the voice of an identifiable natural person otherwise than under the Voice Licensing Addendum;
4. (d)send any message or make any call to a person who has not given the consent required by law and by the applicable Channel policy, or who has opted out; upload a purchased, scraped, rented or third-party contact list; or send unsolicited bulk communications;
5. (e)use a utility or authentication message template, or a message tag, to carry promotional content, or mis-categorise a template;
6. (f)operate the AI Agent as a general-purpose assistant, or for any purpose not ancillary to the Customer’s own goods and services;
7. (g)use the Services to make an automated decision producing legal effects concerning a person or similarly significantly affecting a person; use them in a way that would make them a high-risk or prohibited artificial-intelligence system under applicable artificial-intelligence legislation; or modify their intended purpose so that they could be so classified. Use of the Reliability Score in accordance with Part A of Schedule 5 is not a breach;
8. (h)use the Services for any purpose within clause 10.10;
9. (i)generate or transmit content that is unlawful, defamatory, obscene, harassing, threatening, hateful or discriminatory, or that infringes any intellectual property, privacy or confidentiality right;
10. (j)generate audio or text capable of being mistaken for a recording of, or a statement by, a real event or a real person;
11. (k)transmit malicious code, attempt to gain unauthorised access to the Platform or to any other customer’s data, probe or test the vulnerability of the Platform without Omnistra’s prior written consent, or interfere with the operation, security, integrity or availability of the Platform;
12. (l)impose an unreasonably large load on the Platform through automated means, or exceed a rate limit, quota or usage restriction;
13. (m)submit, or configure the Services to collect, payment card data or data classified as restricted personal data, critical information infrastructure data or national security or defence data under Bangladesh Data Protection Law, each as defined in paragraph 1.1 of Schedule 2; an End Customer volunteering such data is not of itself a breach, but the Customer shall notify Omnistra promptly on becoming aware of it;
14. (n)target, direct, promote or market the Services, or any goods or services offered through them, at persons in the Regulated Markets, or configure an outbound message or call campaign to such persons, without Omnistra’s written agreement under clause 9.5(g); or fail, within 30 days of a notice under that clause that interactions with such persons have become a recurring and material part of the Customer’s use, either to obtain that agreement or to cease that use. An incidental inbound communication initiated by a person in a Regulated Market, about goods or services to be supplied within Bangladesh, is not a breach;
15. (o)breach the acceptable-use or content policy of any Channel or of any Third-Party Dependency;
16. (p)generate, solicit, transmit, store or facilitate any material that sexually exploits or abuses a child, or depicts a minor in a sexualised manner, or attempt to elicit any such material from the AI Agent. Omnistra will report it to law enforcement and each affected Supplier, preserve the records under clause 15.6, and terminate immediately under clause 21.4(a) without the benefit of paragraph 4.2;
17. (q)generate or transmit material that outrages or is intended to outrage the religious feelings of any class of person, that incites hostility, hatred or violence between communities, classes or religions, or that constitutes propaganda for or support of a proscribed organisation or a terrorist act;
18. (r)generate or transmit material that encourages, instructs in or facilitates suicide, self-harm or disordered eating; and the Customer shall configure the AI Agent, on detecting indications of such risk, to escalate to a human and give the referral information Omnistra makes available in the Account;
19. (s)deal in, quote, arrange or facilitate any transaction in cryptocurrency or virtual assets, or any foreign-exchange transaction otherwise than through an authorised dealer, or any transfer of value outside the banking channel;
20. (t)generate malicious code, or generate content designed to deceive as to the identity of a bank, a mobile financial services provider, a government body or a delivery service;
21. (u)direct the Services at persons under 18; and the Customer shall notify Omnistra within seven days of becoming aware that its customer base is likely to include persons under 18, and shall configure the AI Agent to end the conversation and escalate to a human on detecting that the End Customer is under 18;
22. (v)resell, sublicense, rent, lease, or operate the Services as a service bureau or on behalf of any person other than the Customer; or
23. (w)attempt to disable, circumvent, degrade or defeat a Guardrail, a prohibited-claim filter, a protective measure under clause 10.7 or a calling control under clause 11.7, whether by configuration, system instruction, prompt, persona, iterative testing or otherwise; instruct or induce the AI Agent to act inconsistently with any of them; or use the Services to develop, test or refine a technique for doing so, whether against the Platform or against any other artificial-intelligence system.

4. Enforcement

4.1Omnistra may remove or disable offending content or configuration, suspend the Services immediately under clause 14.4, and terminate under clause 21.4(a). It may report a matter to a regulator, a Supplier or law enforcement where required by law or by its agreement with that Supplier.
4.2Where Omnistra proposes to terminate under clause 21.4(a) for a breach which, in its reasonable opinion, has ceased and caused no loss to Omnistra, a third party or an End Customer, it will first notify the Customer, give it five Business Days to explain, and consider any explanation received in that period. Omnistra shall record in writing its reasons for concluding that this paragraph does not apply and provide them to the Customer with the termination notice. A failure to comply does not invalidate a termination but entitles the Customer to a refund of the unused Prepaid Balance despite any deduction under clause 21.6 and to the Exit Refund. This paragraph does not apply to a breach of paragraph 2 or of paragraph 3.1(a), (b), (c), (p), (q) or (r), or where a Supplier, a regulator or law enforcement has required or requested action; and does not restrict Omnistra’s right to suspend immediately.
4.3The Customer’s indemnity for breach of this Policy is in clause 20.2(a). It is uncapped only as clause 19.5(e) provides, and is otherwise capped as clause 19.4(f) provides. This paragraph does not vary clause 19 or clause 20.