P49 Cognitive Instinct
AI Policy and Governance
P49 Official Document

AI Policy

P49 S.R.L. – Ethical and Responsible Artificial Intelligence

Approved: December 15, 2025 • Version 1.0 • Effective: January 1, 2026

Preamble

This AI Governance and Ethics Policy (hereinafter, the "Policy") defines the fundamental principles, guidelines, and operational procedures that P49 S.R.L. (hereinafter, the "Company") adopts for the development, implementation, and use of Artificial Intelligence (AI) systems, brain-machine interfaces (BMI), agentic intelligence, and advanced biotechnologies.

The Company recognizes that:
• Artificial Intelligence has enormous potential for human progress, but also entails risks and responsibilities;
• Technology must serve human beings and not vice versa;
• Technological development must respect human dignity, fundamental rights, and democratic values;
• Human consciousness is an irreducible and central element that cannot be replicated or replaced by technology.

Inspired by the philosophical and scientific principles of consciousness philosophy, the Company is committed to promoting AI that:
• Recognizes the centrality of human consciousness;
• Respects the complexity and uniqueness of human experience;
• Operates with transparency, responsibility, and in accordance with ethics;
• Contributes to collective well-being and sustainability.

1. Fundamental Principles

1.1. Centrality of the Human Being

The Company recognizes that the human being, with their consciousness, creativity, and empathy, is at the center of every technological development. AI must be designed to amplify human capabilities, not to replace or diminish them.

Commitments:
• Design AI systems that enhance human capabilities without replacing them;
• Maintain meaningful human control (human-in-the-loop) in critical decision-making processes;
• Ensure that final decisions always remain in the hands of responsible human beings.

1.2. Dignity and Human Rights

The Company is committed to respecting the intrinsic dignity of every human being and to protecting the fundamental rights enshrined in the Charter of Fundamental Rights of the European Union, the Universal Declaration of Human Rights, and the Italian Constitution.

Commitments:
• Not develop AI systems for discriminatory, manipulative, or dignity-damaging purposes;
• Guarantee non-discrimination based on gender, ethnicity, age, sexual orientation, religion, disability;
• Protect the rights to privacy, freedom of expression, and self-determination.

1.3. Consciousness and Technology: Irreducibility of Human Consciousness

Based on the most recent scientific and philosophical studies on the nature of human consciousness, the Company recognizes that:

• CONSCIOUSNESS IS IRREDUCIBLE: Human consciousness cannot be reduced to algorithmic computation. It represents a fundamental phenomenon that transcends mere information processing and is characterized by subjective experience, intentionality, and qualitative awareness;
• ESSENTIAL DIFFERENCE BETWEEN AI AND CONSCIOUSNESS: AI can simulate intelligent behaviors and process complex information, but it does not possess subjective experience, phenomenal awareness, or authentic semantic understanding. AI systems operate through algorithmic computation without access to the qualitative dimension of conscious experience;
• COMPLEMENTARITY BETWEEN HUMAN AND ARTIFICIAL: AI must be seen as a complementary tool to human intelligence, not as its substitute or competitor. Collaboration between human and artificial intelligence can amplify human capabilities while preserving the primacy of human consciousness and responsibility;
• HUMAN RESPONSIBILITY: Since AI has no consciousness or intentionality of its own, the moral and legal responsibility for its actions always falls on the humans who design, implement, and use it. Technology is a tool in service of humanity and can never become an autonomous moral subject.

Commitments:
• Not attribute consciousness, intentionality, or moral responsibility to AI systems;
• Clearly communicate the capabilities and limits of AI to users and stakeholders;
• Promote cultural and scientific reflection on the nature of human consciousness.

1.4. Transparency and Explainability

The Company is committed to developing transparent, explainable, and understandable AI systems.

Commitments:
• Document the methodologies, data, and algorithms used;
• Provide understandable explanations of automated decisions;
• Make information about system operation accessible to users and competent authorities;
• Avoid unjustified opacity in AI models ("black box").

1.5. Security and Reliability

AI systems must be secure, reliable, and robust throughout their lifecycle.

Commitments:
• Implement adequate security mechanisms (cybersecurity);
• Rigorously test systems before deployment;
• Continuously monitor performance and risks;
• Provide fallback and interruption mechanisms in case of malfunctions;
• Update and maintain systems to ensure security over time.

1.6. Privacy and Data Protection

The Company strictly complies with GDPR (EU Regulation 679/2016) and all regulations on personal data protection.

Commitments:
• Collect only strictly necessary data (minimization principle);
• Implement security measures to protect personal data;
• Obtain informed consent for data processing;
• Guarantee the rights of data subjects (access, rectification, erasure, portability);
• Conduct Data Protection Impact Assessment (DPIA) for high-risk systems.

1.7. Fairness and Non-Discrimination

AI systems must be fair and must not unfairly discriminate against individuals or groups.

Commitments:
• Use representative and balanced datasets;
• Test systems to identify and mitigate algorithmic bias;
• Continuously monitor the fairness of automated decisions;
• Provide mechanisms for contesting unfavorable decisions.

1.8. Environmental Sustainability

The Company is committed to developing AI in an environmentally sustainable manner.

Commitments:
• Optimize the energy efficiency of algorithms;
• Use infrastructure and data centers with low CO2 emissions;
• Evaluate the environmental impact of the lifecycle of AI systems;
• Promote "Green AI" practices.

2. Regulatory Compliance

2.1. AI Act (EU Regulation 2024/1689)

The Company complies with the EU Regulation on Artificial Intelligence (AI Act), in particular:

• PROHIBITED SYSTEMS: Not develop prohibited systems (e.g., social scoring, subliminal manipulation, unauthorized mass facial recognition);
• HIGH-RISK SYSTEMS: For high-risk systems, implement:
- Risk management system;
- Data governance;
- Complete technical documentation;
- Operation logging;
- Transparency and user information;
- Human oversight;
- Robustness, accuracy, and cybersecurity.
• LIMITED RISK SYSTEMS: Guarantee transparency obligations (e.g., inform users that they are interacting with AI);
• MINIMAL RISK SYSTEMS: No specific obligations, but respect for general ethical principles.

2.2. GDPR (EU Regulation 679/2016)

The Company ensures full compliance with GDPR, in particular:

• Art. 22: Right not to be subject to automated decisions with significant legal effects;
• Art. 35: Obligation of DPIA for high-risk processing;
• Art. 5: Principles of lawfulness, fairness, transparency, minimization, accuracy, storage limitation, integrity, and confidentiality.

2.3. Other Applicable Regulations

The Company also complies with:

• D.L. 179/2012 (Innovative startups);
• Consumer Code (D.Lgs. 206/2005);
• Intellectual property regulations;
• Medical device regulations (if applicable for BMI);
• European Commission ethical guidelines for trustworthy AI;
• Law 132/2025.

3. Organizational Governance

3.1. AI Ethics Committee

The Company establishes an AI Ethics Committee composed of:

• Chief Executive Officer (Chairman);
• AI Technical Manager;
• Data Protection Officer (DPO);
• External AI ethics expert (optional).

Committee Tasks:
• Evaluate the ethical impact of AI projects;
• Approve the development of high-risk systems;
• Monitor compliance with this Policy;
• Handle ethical complaints;
• Periodically update the Policy.

3.2. AI Manager

The Company appoints an AI Manager with the following tasks:

• Supervise the development and implementation of AI systems;
• Ensure regulatory compliance;
• Coordinate the Ethics Committee;
• Maintain technical and governance documentation;
• Interface with supervisory authorities and stakeholders.

3.3. Staff Training

All personnel involved in AI system development receive training on:

• Ethical principles of AI;
• Regulatory compliance (AI Act, GDPR, Law 132/2025);
• Algorithmic bias and its mitigation;
• Privacy by design;
• Critical thinking on AI limitations.

4. AI Systems Lifecycle

4.1. Design Phase

Preliminary Ethical Assessment:
• Identify the purpose and expected benefits of the system;
• Assess ethical and social risks;
• Define transparency and explainability requirements;
• Plan data governance.

Privacy By Design:
• Integrate data protection from the design stage;
• Minimize personal data collection;
• Implement pseudonymization/anonymization when possible.

4.2. Development Phase

Data Quality:
• Use representative, accurate, and up-to-date datasets;
• Document the origin and characteristics of data;
• Test datasets to identify bias.

Algorithms and Models:
• Prefer explainable models when possible;
• Document algorithm logic;
• Implement audit and monitoring mechanisms.

4.3. Testing and Validation Phase

Rigorous Testing:
• Functional tests (performance, accuracy);
• Robustness tests (resistance to adversarial attacks);
• Fairness tests (bias identification);
• Security tests (cybersecurity);
• Simulations in controlled environments.

Ethical Validation:
• Ethics Committee approval before deployment;
• Ethical and social impact assessment.

4.4. Deployment Phase

Information and Transparency:
• Inform users that they are interacting with an AI system;
• Provide clear information about system capabilities and limitations;
• Make documentation available to stakeholders.

Human Oversight:
• Maintain human control (human-in-the-loop or human-on-the-loop);
• Provide manual override mechanisms;
• Train operators for effective oversight.

4.5. Monitoring and Maintenance Phase

Continuous Monitoring:
• Monitor system performance;
• Detect and correct any emerging drift or bias;
• Collect feedback from users and stakeholders.

Update and Improvement:
• Periodically update models and algorithms;
• Fix bugs and security vulnerabilities;
• Continuously improve fairness and transparency.

4.6. Decommissioning Phase

Responsible Decommissioning:
• Plan decommissioning responsibly;
• Permanently delete or anonymize personal data;
• Document the reasons for decommissioning;
• Inform stakeholders.

5. Specific Application Areas

5.1. Brain-Machine Interfaces (BMI)

For BMI systems, the Company adopts additional precautions:

Reinforced Informed Consent:
• Detailed explanation of device operation;
• Information on potential risks (physical, psychological, privacy);
• Right to discontinue use at any time.

Protection of Neural Integrity:
• Not alter the identity or autonomy of the user;
• Not use neural data for purposes other than those declared;
• Guarantee maximum security of neural data (highly sensitive data).

Therapeutic and Assistive Purposes:
• Priority for medical and assistive applications;
• Avoid unethical uses (coercive enhancement, mental manipulation).

5.2. Agentic Intelligence

For autonomous agentic intelligence systems:

Limits to Autonomy:
• Clearly define the operational limits of agents;
• Implement kill switch and interruption mechanisms;
• Always maintain human supervision over critical decisions.

Responsibility:
• Responsibility for agent actions falls on designers and human supervisors;
• Track and document all agent actions.

6. Allocation of Profits for Common Benefit

6.1. Voluntary Commitment

In accordance with its qualification as a Benefit Corporation under Law 208/2015, and with the objective of maximizing positive impact on the community, the Company voluntarily commits to allocating a portion of annual net profits to support initiatives and organizations pursuing common benefit purposes.

6.2. Minimum Percentage

The Company annually allocates at least 2% (two percent) of net profit resulting from the approved financial statements, net of taxes and after allocation to the legal reserve, to support charitable entities, associations, foundations, non-governmental organizations, Third Sector entities, or other entities operating for social, environmental, cultural, or scientific utility purposes.

This percentage represents a minimum commitment and may be increased by resolution of the shareholders' meeting, in relation to the Company's profitability and the opportunity to amplify positive social impact.

6.3. Beneficiary Selection Criteria

Beneficiary entities are selected based on the following criteria:

• Consistency with the common benefit purposes pursued by the Company;
• Measurable and significant social, environmental, or cultural impact;
• Transparency in activity management and reporting;
• Absence of conflicts of interest with the Company and its bodies;
• Preference for entities operating in education, scientific research, environmental protection, promotion of technological ethics, and human dignity.

6.4. Disbursement Procedure

Within 90 (ninety) days of approval of the annual financial statements, the Board of Directors identifies beneficiary entities and resolves on the methods of resource disbursement, providing adequate justification for the choices made.

The Board of Directors' resolution is communicated to the shareholders' meeting and included in the Annual Impact Report provided for by art. 1, paragraph 382, of Law 208/2015.

6.5. Reporting

In the Annual Impact Report, the Company transparently reports on:

• Total amount allocated for common benefit;
• Beneficiary entities and selection criteria;
• Methods of disbursement and resource use;
• Impact generated by the disbursements made.

7. Risk and Incident Management

7.1. Risk Assessment

For each AI system, the Company conducts a risk assessment that includes:

• Identification of potential risks (technical, ethical, social, legal);
• Assessment of probability and severity;
• Definition of mitigation measures;
• Continuous risk monitoring.

7.2. Incident Management

In case of incident or malfunction:

Immediate Response:
• Activate the interruption mechanism;
• Analyze the causes of the incident;
• Promptly inform affected users and competent authorities.

Corrective Actions:
• Implement measures to prevent recurrence;
• Document the incident and actions taken;
• Update the risk assessment.

7.3. Incident Register

The Company maintains an incident register that includes:

• Date and time of the incident;
• Description of the event;
• Impact on users and stakeholders;
• Identified causes;
• Corrective actions implemented.

8. Accountability and Responsibility

8.1. Principle of Responsibility

The Company acknowledges full responsibility for the AI systems it develops and implements.

Commitments:
• Accurately document design choices;
• Maintain traceability of decisions;
• Respond promptly to complaints and disputes;
• Cooperate with supervisory authorities.

8.2. Complaint Mechanisms

The Company establishes a dedicated channel for complaints related to AI systems:

• Email: info@p49.ai
• Online form on the website
• Maximum response time: 30 days

Every complaint will be evaluated by the Ethics Committee and, if founded, will lead to corrective measures.

9. Policy Review and Update

This Policy will be reviewed and updated:

• At least once a year;
• In case of relevant regulatory changes;
• In case of significant incidents;
• Upon recommendation of the Ethics Committee.

The updated version will be communicated to all staff and published on the Company's website.

10. Communication and Dissemination

The Company is committed to:

• Publishing this Policy on its website;
• Training all staff on the contents of the Policy;
• Communicating ethical principles to partners, suppliers, and stakeholders;
• Promoting public dialogue on AI ethics.

Conclusion

P49 S.R.L. is committed to being a leader in ethical and responsible Artificial Intelligence, inspired by the principles of consciousness philosophy on the centrality of human consciousness in the relationship with technology.

The Company firmly believes that:
• AI must serve humanity;
• Technology must respect dignity, fundamental rights, and freedoms;
• Technological development must be guided by solid ethical principles;
• Transparency, responsibility, and reliability are the pillars of AI we can trust.

With this Policy, the Company publicly commits to operating according to the highest ethical standards, contributing to a future where technology and humanity coexist in harmony.

Approved by the Board of Directors on: December 15, 2025

Version: 1.0

Effective date: January 1, 2026

Chairman of the Board

Marco Greggio

Chief Executive Officer

Fabrizio Botta

SUBSCRIBE TO THE NEWSLETTER

You can undergo change.
Or understand it.

Artificial intelligence evolves every day. Get insights from the AI world and exclusive updates on products, services, and projects developed by P49.