Policy

acceptable-use-policy

Acceptable Use Policy

AlphaGen Holdings Limited ("AlphaGen")

Companies House no. 17084844 — registered in England and Wales

Effective date: 2026-04-27

Version: 1.0.0

Contact: abuse@alpha-gen.ai (incidents) · legal@alpha-gen.ai

(policy questions)

This Acceptable Use Policy (the "AUP") governs your use of

the AlphaGen AutoAnnotation System, the AlphaGen mobile

application, our developer APIs, and any related software or

data products (collectively, the "Services"). The AUP is

incorporated into the Terms of Service

and applies to every user, including each Authorised User

operating under a Customer account.

If you breach this AUP, AlphaGen may suspend or terminate your

access under Section 11 of the Terms of Service, with or without

prior notice depending on severity, and may report unlawful

activity to law-enforcement authorities.

---

1. Lawful use

You must use the Services only for lawful purposes. You must

not use the Services to:

Violate any applicable law, regulation, court order, or

binding regulator guidance — including UK and EU

data-protection laws, sanctions, export controls, anti-money-

laundering rules, or rules of any professional body governing

your sector.

Infringe any third party's intellectual property, privacy,

publicity, or contractual rights.

Process personal data without a valid legal basis under UK

GDPR / EU GDPR or equivalent law.

2. Prohibited content

You must not upload, store, transmit, generate, or annotate any

content that:

Depicts child sexual abuse material (CSAM) or any sexual

exploitation of minors. Suspected CSAM is reported to the

Internet Watch Foundation (IWF) and to the National Center for

Missing & Exploited Children (NCMEC) where applicable, with

account suspension and full preservation of evidence under

s.69 Serious Crime Act 2015 procedures.

Depicts non-consensual intimate imagery, "deepfake" sexual

content, or other content that violates the Online Safety Act

2023.

Constitutes incitement to terrorism, glorification of violence,

or recruitment material under the Terrorism Acts.

Is defamatory, libellous, threatening, harassing, or

intimidating.

Discloses personal data of a third party without that person's

documented consent or another lawful basis.

Contains malware, exploit code, or material designed to

compromise the Services or other systems.

3. Prohibited uses

Even where the underlying content is lawful, you must not use

the Services to:

3.1 Surveillance and identification

Conduct mass surveillance of identifiable individuals not

involved in your specified Customer use case.

Build, train, or operate any face-recognition, gait-recognition,

or biometric identification system targeting people who have

not provided documented consent for that purpose, except where

expressly contracted under an Order Form that names the

permitted use.

Re-identify any subject whose identifiers have been suppressed

or redacted by the Services.

3.2 Profiling and decision-making

Use Output to make automated decisions producing legal or

similarly significant effects on a person under Art. 22 UK GDPR

unless you have a valid Art. 22(2) basis and have implemented

the safeguards required by law.

Use Output to score, profile, or rank natural persons for

insurance, credit, employment, housing, education,

immigration, benefits, or law-enforcement purposes without

written authorisation under your Order Form.

3.3 Health, legal, financial, safety-critical

Rely on Output as the sole basis for any clinical diagnosis,

treatment decision, legal opinion, financial advice, or

safety-critical determination. AlphaGen Output is

diagnostic-grade only when expressly certified for that use on

the Order Form, and even then, qualified human review is

required.

3.4 Manipulation and deception

Generate, alter, or label content with intent to deceive,

including fabricating evidence, impersonating real people, or

misrepresenting AI-generated content as human-authored where

the audience would reasonably expect human authorship (see the

AI Transparency page).

3.5 Competitive use

Use the Services to develop, train, or benchmark a competing

AI / annotation product, or to extract AlphaGen models, model

weights, prompts, or training data.

Reverse-engineer, decompile, or otherwise attempt to derive

the source code, model weights, or proprietary algorithms of

the Services, except to the extent expressly permitted by

law.

4. System integrity

You must not:

Probe, scan, or test the vulnerability of the Services, except

through AlphaGen's authorised vulnerability-disclosure process

(security@alpha-gen.ai).

Bypass, disable, or interfere with any rate-limiting,

authentication, authorisation, or content-filtering control.

Circumvent any usage limit on your Order Form (e.g.

programmatically rotating accounts to evade per-account quotas).

Generate or distribute spam, mass unsolicited messages, or

conduct phishing through the Services.

Use the Services to mine cryptocurrency, host or relay

network attacks, run torrent trackers, or conduct any

resource-intensive activity outside the contracted purpose.

Submit data with the intent of exhausting infrastructure,

causing denial of service, or otherwise harming AlphaGen,

other customers, or third parties.

5. AI-specific prohibitions

You must not use the Services to:

Train weapons systems, military targeting systems, or

autonomous lethal-force decision systems.

Generate content for political deception — e.g. fabricated

speech audio of real political figures, fake election

announcements, or content designed to suppress voter turnout.

Generate non-consensual sexual content of any real person.
Defeat copyright by generating content substantially

reproducing a specific copyrighted work the user does not have

rights to.

Generate harmful instructions — e.g. instructions for

manufacturing weapons of mass destruction, biothreats,

malware, or for self-harm.

Bypass another platform's safety systems (jailbreak prompts,

payload obfuscation).

These prohibitions apply regardless of how the Service is being

used — annotation, propagation, model fine-tuning (LoRA), or

analytics workflows.

6. HITL game integrity

When using the AlphaGen mobile HITL games (Masking, Salience,

Intent, Discovery), you must not:

Submit deliberately false or low-effort annotations to claim

game points or leaderboard position.

Operate multiple accounts to inflate your own performance

metrics or to manipulate trust-weighting calculations.

Share screen recordings, prompts, or accounts with people who

do not hold an Authorised User credential under the Customer's

account.

Reverse-engineer the trust-weighting model, the conductor's

selector, or the cognitive-load model.

7. Reporting abuse

If you become aware of any breach of this AUP — by your own

Authorised Users or by a third party — please notify

abuse@alpha-gen.ai promptly. AlphaGen treats abuse reports as

confidential and will not retaliate against good-faith reporters.

8. Enforcement

Suspected breaches of this AUP are investigated under the

Internal Information Security Policy by AlphaGen's Trust &

Safety on-call. The investigation may include reviewing relevant

audit logs, blocking specific API endpoints or IP ranges,

suspending an Authorised User credential, or preserving evidence

under legal hold. AlphaGen will give the Customer reasonable

notice of any enforcement action where lawful and operationally

practicable.

The remedies for breach of this AUP are cumulative with those in

the Terms of Service and at law.

9. Updates

AlphaGen may update this AUP from time to time. The current

version is always at docs/legal/external/acceptable-use-policy.md

and on https://alpha-gen.ai/legal/acceptable-use-policy. Material

changes are announced in the Privacy Policy Changelog and (for

contracted Customers) by email to the account contact.

---

Document control

|---|---|---|---|

| 1.0.0 | 2026-04-27 | AlphaGen Legal | Initial publication. |