AIFORYA Ethics Charter: Truth & Corporate Anonymity

Preamble

Artificial intelligence is no longer a distant promise. It is an active force deeply transforming professions, businesses, and societies. Faced with this reality, inaction is not an option.

AIFORYA, as a designer of software tools augmented by artificial intelligence, assumes its responsibility. This responsibility demands making clear choices, setting firm limits, and building on solid moral foundations.

This document is neither a communication exercise nor a collection of empty wishes. It is a reference framework — the ethical operating system governing AIFORYA's decisions. It defines, in a public and transparent manner, the principles guiding the development of every product, every line of code, and every strategic decision.

This charter is a public, evolving, and auditable commitment. It sets out what AIFORYA does — and more importantly, what AIFORYA refuses to do.

Commitments in Brief

Total client sovereignty: through the BYOK (Bring Your Own Key) model, AI data and costs remain under the client's exclusive control. AIFORYA is never an intermediary.
Radical transparency: no hidden margins, no opaque credits. The client pays the real cost of AI and understands how the tools work.
GDPR+ data protection: confidentiality is embedded in the very architecture of products. AIFORYA collects only what is essential.
Absolute prohibitions: AIFORYA will never develop technologies for autonomous weaponry, mass surveillance, organised disinformation, or cognitive manipulation.
Guaranteed independence: the independent ownership structure protects AIFORYA from external pressures that could compromise its ethical principles.
Inflexible quality: AIFORYA commits to providing reliable, performant, and maintained tools — because quality is a form of respect.

Part 1 — What AIFORYA Does: Core Principles in Action

1.1. Augmentation Tools, Not Replacement

AIFORYA designs co-pilots, not autopilots. Software is built to augment human capabilities — the developer's creativity, the marketer's insight, the customer service agent's efficiency.

The goal is never to replace humans, but to free their potential from repetitive tasks so they can focus on strategy, critical judgement, and creativity.

1.2. Client Sovereignty via the BYOK Model

AIFORYA operates on a fundamental model: Bring Your Own Key. Each client uses their own API key (Anthropic, OpenAI, Google...) to power the AI features in extensions.

Total data control: data sent to AI models transits directly between the client's infrastructure and their chosen AI provider.
Absolute cost control: the client pays their AI provider directly for consumption. No AIFORYA margin on AI.
No lock-in: usage history belongs to the client and remains with their AI provider.

1.3. Maximum Transparency on Costs and Operation

AI costs: interfaces display a clear cost estimate for each AI operation, based on public provider pricing grids.
Models used: AI model choice is left to the client where technically relevant.
Detailed changelogs: every update is accompanied by a public change log.

1.4. Data Protection Beyond Standards (GDPR+)

Minimisation: AIFORYA collects only what is strictly necessary (licence management, billing).
Anonymisation by default: aggregated usage data is fully anonymised.
Security: in-transit and at-rest encryption, environment isolation, proactive monitoring.

1.5. Quality as Non-Negotiable Standard

Reliability: automated unit, integration, and regression tests.
Performance: obsessive profiling of every feature.
Responsive technical support: included in the offer, no premium-tier gating.

1.6. Responsible and Honest Development

AIFORYA designs clear, honest interfaces that respect the user's intelligence and time. Design serves functionality and autonomy — never manipulation.

Part 2 — What AIFORYA Never Does: Absolute Red Lines

2.1. No Autonomous Weapons

AIFORYA will never participate, directly or indirectly, in the design, development, or commercialisation of lethal autonomous weapons systems.

2.2. No Pathogens or Dangerous Biological Tools

AIFORYA will not use its technologies to design, model, or facilitate the creation of pathogenic agents or biological weapons.

2.3. No Non-Therapeutic Human Genetic Manipulation

AIFORYA will not develop tools aimed at facilitating manipulation of the human germline for non-therapeutic enhancement purposes.

2.4. No Mass Individual Surveillance

AIFORYA refuses to build systems designed for mass surveillance, non-consensual individual identification, systematic tracking, or social scoring.

2.5. No Dark Patterns or Cognitive Manipulation

AIFORYA prohibits integrating dark patterns: false urgency, artificial scarcity, complex unsubscription flows, etc.

2.6. No Organised Disinformation

AIFORYA will not create tools whose primary purpose is generating content for disinformation operations, targeted political propaganda, or content farms.

2.7. No Sale of Personal Data

AIFORYA will never sell, rent, or transfer clients' personal data to third parties for commercial purposes.

2.8. No Partnerships with Authoritarian Regimes

AIFORYA prohibits entering into bespoke development contracts with entities whose manifest purpose is political repression or violation of fundamental human rights.

2.9. No Addictive Technologies by Design

AIFORYA will not design products aimed at maximising engagement time through exploitation of cognitive biases leading to dependency.

2.10. No Intentional Generation of Illegal Content

AIFORYA prohibits designing systems whose explicit purpose is to circumvent AI model safety filters to generate illegal content.

Part 3 — Ethical Governance

Technology retention: AIFORYA reserves the right not to distribute certain technologies if their misuse potential is deemed too high.
Internal auditability: strategic decisions traced and documented.
External transparency: charter, T&Cs, privacy policy, and legal notices publicly accessible.
Reversibility: via BYOK, the client can cease using services at any time without losing their AI assets.
Acknowledged fallibility: in the event of error, AIFORYA commits to recognising it, correcting course, and evolving this charter.
Independence: AIFORYA is an independent French company. Decisions are guided by values and clients' long-term interests.

Part 4 — The BYOK Model as Structural Ethical Commitment

| Classic SaaS Trap | BYOK Solution | | :--- | :--- | | Data stored with the publisher — risk of leak or re-use. | Direct client → AI provider flow. By design, AIFORYA has no access to confidential data. | | Opaque credit-based pricing masking real cost. | Client billed directly by their AI provider at public rates. AIFORYA has no interest in driving up consumption. | | Lock-in via difficult history export. | History stored with the client's provider. Full freedom to migrate. | | Temptation to degrade entry-level offers to force upgrades. | AI quality = quality of model chosen by client. AIFORYA cannot throttle the output. |

Part 5 — Concrete and Measurable Commitments

Service availability: objective above 99.5% annually for critical AIFORYA-managed services.
Technical support: first qualified response within 2 business days maximum.
Right of withdrawal: full refund without conditions within 14 days for any new purchase.
Update transparency: systematic publication of a public, detailed changelog.
Incident transparency: public communication within 48 hours in the event of a major incident.
Charter review: at minimum once per year.
Accessibility: continuous effort towards WCAG 2.1 Level AA.

Part 6 — Contact and Ethics Complaint

For any question, concern, or report regarding application of this charter, a dedicated address is available:

ethique@aiforya.fr

Guaranteed complaints process:

Receipt: acknowledgement within 2 business days.
Analysis: impartial internal review.
Response: substantiated response within 15 business days.
Action: immediate corrective measures if a breach is established.

If disagreement persists following the internal process, AIFORYA accepts recourse to an independent external mediator.

Part 7 — Evolution

This document is a snapshot of AIFORYA's commitments at its publication date. As the AI world evolves continuously, this charter is a living document: reassessed at minimum annually, or more frequently if major technological advances or regulatory developments warrant it.

Community feedback and suggestions are actively encouraged — they contribute to strengthening the relevance and robustness of these commitments.

This charter is more than a document: it is AIFORYA's trust contract with the world.

Version 1.0 — AIFORYA, independent French company.