Playbook
Generative AIat Asylum
v 2.0
As GenAI rapidly reconfigures the landscape
of advertising and marketing, its real value
is found in the integration of robust frameworks:
human-in-the-loop control, risk mitigation,
and end-to-end traceability.
Generative AI allows us to accelerate ideation and content production across multiple formats, reducing repetitive tasks and enabling more creative iterations in less time. Its greatest value emerges when integrated with clear processes: human oversight, brand consistency, and workflow traceability.
With this approach, we can adapt messages and variants by channel or audience more efficiently, always under a responsible use of information and prior validation before publishing or delivering to clients.
We seek to create more relevant experiences while strengthening our teams’ capabilities: AI as a tool to elevate speed and quality, not to replace professional judgment.
This playbook defines how we use generative AI at Asylum: best practices, security and privacy criteria, transparency
where applicable, and a risk-based work approach centered on human oversight and traceability.
Generative AI must be applied responsibly: respecting privacy, rights, and intellectual property, while protecting the information of our clients and audiences. These guidelines are cross-functional and apply to all teams and processes, especially for any asset that is published or delivered to clients. Therefore, we require human oversight before disclosure, the use of approved tools, and minimum process traceability.
BENEFITS
Inspiration
& Innovation
Generative AI expands creative exploration and accelerates the generation of pathways, references, and variations, elevating the quality of proposals and executions through creative direction and human oversight.
Relevance & Responsible Personalization
It enables the adaptation of messages and formats by context, channel, or segment, increasing relevance without compromising privacy or brand consistency.
Efficiency
& Productivity
Reduces repetitive tasks and accelerates definition, ideation, and production, enabling more iterations in less time. The output is validated through QA before publishing or delivery to clients.
Continuous Improvement through Operational Learning
We improve with use: we record what works, standardize templates, and optimize workflows, increasing consistency and quality over time.
BEST
PRACTICES
FOR THE USE OF GENAI
TRANSPARENCY & DISCLOSURE
When the use of AI may create confusion regarding authenticity, identity, testimonials, or the nature of the content, we declare it clearly and proportionally. Transparency is applied through a risk-based approach: not everything requires the same level of notice, but nothing should be misleading.
DATA PROTECTION & SECURITY
We use AI while respecting privacy and confidentiality: we practice data minimization, avoid exposing sensitive information, and work exclusively with agency-approved tools and accounts. Client and audience data are managed under strict internal controls.
VALIDATION & QA BEFORE PUBLISHING OR DELIVERY
Before publishing or delivering to a client, all AI-generated content undergoes QA to ensure: creative quality, brand consistency, accuracy of claims, rights infringement risks (intellectual property), and case-specific compliance.
HUMAN OVERSIGHT & TRACEABILITY
AI assists; the team decides. No AI-generated content is published or delivered to clients without human review and approval. Furthermore, we maintain a minimum process record (tool/model used, relevant inputs, editing decisions, and final version) to ensure consistency, accountability, and continuous improvement.
GENERAL
RECOMMENDATIONS
TOOLS & TECHNOLOGIES
We evaluate and approve AI tools based on their performance and, above all, their commercial viability: security, data handling, licensing, and compatibility with our needs and those of each client.
PROCESSES & METHODOLOGIES
We review workflows to identify where AI generates real value—speed, quality, or scale—and we define the required level of oversight based on the type of asset, the target audience, and the specific risk of each case.
BRAND GUARDIANS
We ensure creative consistency and brand protection: correct use of names, products, and messaging; control of claims; visual and verbal cohesion; and mandatory validation before publishing or delivery.
ADOPTION
We drive adoption through training, shared templates, and unified QA criteria. We support teams in implementing approved workflows and ensuring continuous improvement based on operational learning.
RULES &
CONSIDERATIONS
ETHICS
Bias & Discrimination
AI can reflect biases present in data and cultural patterns. We apply human review to detect and correct stereotypes, discrimination, or exclusion, especially regarding gender representation, ethnicity, age, body types, disability, and sensitive contexts.
Truth, Authenticity & Non-Deception
We do not use AI to mislead regarding facts, testimonials, results, or identities. Any representation that could be confused with real material (e.g., people, voices, or situations) requires strict validation and, where appropriate, disclosure proportional to the risk.
Consent & Privacy
We only use personal data when there is an authorized basis and proper permissions. We minimize the use of sensitive data and avoid exposing client, audience, or project information in unapproved tools.
Human Oversight & Generated Content QA
No AI-generated asset is published or delivered without human review and approval. The responsible team ensures brand consistency, reputational safety, and compliance before distribution.
Licensing, Rights & Corporate Subscriptions
We use AI tools under licenses compatible with commercial use and client deliverables. Subscriptions and accounts must be corporate (company domain), avoiding personal emails to guarantee access control, security, operational continuity, and compliance with legal and data protection criteria.
UPDATES
& ADAPTATION
GenAI technology is constantly evolving; therefore, it is essential to stay updated on the latest developments, research, and best practices in the field.
Furthermore, it is necessary to adapt this playbook to the methodological, technical, regulatory, and legal changes specific to each country and market where we operate.
For this reason, this Playbook is subject to constant changes and modifications, which will be incorporated and updated within this document.
© 2026 Asylum Marketing – For INTERNAL use only.
Playbook
Generative AIat Asylum
v 2.0
As GenAI rapidly reconfigures the landscape of advertising and marketing, its real value
is found in the integration of robust frameworks: human-in-the-loop control, risk mitigation, and end-to-end traceability.
Generative AI allows us to accelerate ideation and content production across multiple formats, reducing repetitive tasks and enabling more creative iterations in less time. Its greatest value emerges when integrated with clear processes: human oversight, brand consistency, and workflow traceability.
With this approach, we can adapt messages and variants by channel or audience more efficiently, always under a responsible use of information and prior validation before publishing or delivering to clients.
We seek to create more relevant experiences while strengthening our teams’ capabilities: AI as a tool to elevate speed and quality, not to replace professional judgment.
This playbook defines how we use generative AI at Asylum: best practices, security and privacy criteria, transparency where applicable, and a risk-based work approach centered on human oversight and traceability.
Generative AI must be applied responsibly: respecting privacy, rights, and intellectual property, while protecting the information of our clients and audiences. These guidelines are cross-functional and apply to all teams and processes, especially for any asset that is published or delivered to clients. Therefore, we require human oversight before disclosure, the use of approved tools, and minimum process traceability.
BENEFITS
Inspiration
& Innovation
Generative AI expands creative exploration and accelerates the generation of pathways, references, and variations, elevating the quality of proposals and executions through creative direction and human oversight.
Relevance & Responsible Personalization
It enables the adaptation of messages and formats by context, channel, or segment, increasing relevance without compromising privacy or brand consistency.
Efficiency
& Productivity
Reduces repetitive tasks and accelerates definition, ideation, and production, enabling more iterations in less time. The output is validated through QA before publishing or delivery to clients.
Continuous Improvement through Operational Learning
We improve with use: we record what works, standardize templates, and optimize workflows, increasing consistency and quality over time.
BEST
PRACTICES
FOR THE USE OF GENAI
TRANSPARENCY & DISCLOSURE
When the use of AI may create confusion regarding authenticity, identity, testimonials, or the nature of the content, we declare it clearly and proportionally. Transparency is applied through a risk-based approach: not everything requires the same level of notice, but nothing should be misleading.
DATA PROTECTION & SECURITY
We use AI while respecting privacy and confidentiality: we practice data minimization, avoid exposing sensitive information, and work exclusively with agency-approved tools and accounts. Client and audience data are managed under strict internal controls.
VALIDATION & QA BEFORE PUBLISHING OR DELIVERY
Before publishing or delivering to a client, all AI-generated content undergoes QA to ensure: creative quality, brand consistency, accuracy of claims, rights infringement risks (intellectual property), and case-specific compliance.
HUMAN OVERSIGHT & TRACEABILITY
AI assists; the team decides. No AI-generated content is published or delivered to clients without human review and approval. Furthermore, we maintain a minimum process record (tool/model used, relevant inputs, editing decisions, and final version) to ensure consistency, accountability, and continuous improvement.
GENERAL RECOMMENDATIONS
TOOLS & TECHNOLOGIES
We evaluate and approve AI tools based on their performance and, above all, their commercial viability: security, data handling, licensing, and compatibility with our needs and those of each client.
PROCESSES & METHODOLOGIES
We review workflows to identify where AI generates real value—speed, quality, or scale—and we define the required level of oversight based on the type of asset, the target audience, and the specific risk of each case.
BRAND GUARDIANS
We ensure creative consistency and brand protection: correct use of names, products, and messaging; control of claims; visual and verbal cohesion; and mandatory validation before publishing or delivery.
ADOPTION
We drive adoption through training, shared templates, and unified QA criteria. We support teams in implementing approved workflows and ensuring continuous improvement based on operational learning.
RULES &
CONSIDERATIONS
ETHICS
Bias & Discrimination
AI can reflect biases present in data and cultural patterns. We apply human review to detect and correct stereotypes, discrimination, or exclusion, especially regarding gender representation, ethnicity, age, body types, disability, and sensitive contexts.
Truth, Authenticity & Non-Deception
We do not use AI to mislead regarding facts, testimonials, results, or identities. Any representation that could be confused with real material (e.g., people, voices, or situations) requires strict validation and, where appropriate, disclosure proportional to the risk.
Consent & Privacy
We only use personal data when there is an authorized basis and proper permissions. We minimize the use of sensitive data and avoid exposing client, audience, or project information in unapproved tools.
Human Oversight & Generated Content QA
No AI-generated asset is published or delivered without human review and approval. The responsible team ensures brand consistency, reputational safety, and compliance before distribution.
Licensing, Rights & Corporate Subscriptions
We use AI tools under licenses compatible with commercial use and client deliverables. Subscriptions and accounts must be corporate (company domain), avoiding personal emails to guarantee access control, security, operational continuity, and compliance with legal and data protection criteria.
UPDATING & ADAPTATION
GenAI technology is constantly evolving; therefore, it is essential to stay updated on the latest developments, research, and best practices in the field.
Furthermore, it is necessary to adapt this playbook to the methodological, technical, regulatory, and legal changes specific to each country and market where we operate.
For this reason, this Playbook is subject to constant changes and modifications, which will be incorporated and updated within this document.
© 2026 Asylum Marketing – For INTERNAL use only.