Principles for Responsible AI in Advertising
By Hannah Pavalow, ThinkMedium Client Advisor
As discussed in our piece on Conscious AI Innovation, AI’s proliferation within advertising dictates a need for increased responsibility for its deployment. With ad delivery systems, the creative process, and media insights increasingly being enhanced by AI, our industry, like others, is diving in head-first.
The advantage of all industries simultaneously leaning into a new technology is that it enables learning from each other. While there have been some industry efforts to define AI principles, this post aims to apply industry-agnostic AI governance principles identified by academics to advertising use cases and prompt an advertising industry discussion on defining/refining the responsible usage of AI.
A recent academic article reviewed 200 AI guidelines from across the globe to examine the convergence of principles and trends, identifying 17 distinct principles. Reviewing each generic principle and considering whether it applies to AI use in advertising resulted in an advertising-specific version (full table below). Some of the generic AI principles led to redundant advertising AI principles, producing a final list of 10 principles for responsible AI in advertising.
10 Principles for Responsible AI in Advertising
- Advertising systems that utilize AI must be compliant with the regulatory bodies in the regions they’re deployed in.
- Advertising systems that use AI ought to respect intellectual property rights.
- Advertising systems that use AI ought to uphold individual privacy.
- Advertising systems that utilize AI ought to disclose its usage to users in readily understandable terms.
- Advertising systems that use AI ought to offer an option to enable any user to opt-out of its use.
- Advertising systems that use AI should be evaluated for safety/trustworthiness.
- AI that is deployed within the advertising ecosystem should aim to adhere to “truth in advertising” practices.
- AI that is deployed within the advertising ecosystem should strive towards efficiency.
- Advertising systems that utilize AI ought to explicitly consider diversity and ensure that the outcomes are inclusive.
- The use of AI in advertising should be limited for sensitive verticals and audiences, specifically when the target is children or the topics are politics and health.
These principles focus on compliance, users’ experiences with ad experiences informed by AI, goals of deploying ad systems utilizing AI, and explicit points for intervention and limitation. These four themes provide a framework for thinking about responsible AI in advertising and how to deploy the technology while preserving safety (industry and consumer).
Compliance
These principles are the least contentious, advocating for advertising systems to comply with local regulations and respect IP and privacy. Respecting IP is most relevant in using generative AI in the creative process. Current issues arise when brands use music or creative without artists’ permission and potential harm with AI “borrowing” music, art, or even people’s likenesses when creating a new ad.
Users’ Experiences with Ad Experiences Informed by AI
These principles are most similar to industry trends towards informing users about how their data is used. In many cases where individual data is used in advertising (e.g., personalization), AI is already being used to execute. Hence, the principles of disclosure, opt-out, and safety are not new, but more work needs to be done to extend them to AI.
Deploying AI in Advertising
These principles are likely the most controversial, as the optimization goals and parameters used when creating AI models have the greatest impact on the outcomes of those models. Building truthfulness, efficiency, and diversity into the advertising ecosystem could improve the overall digital advertising ecosystem and its impact. In some advertising verticals, diversity is already explicitly considered; for example, housing ads cannot include ethnicity targeting. Truthfulness is philosophically covered by “false advertising”, but the specific application within AI would need to be refined. Efficiency is currently not legislatively endorsed, but Made For Advertising sites, which often rely on GenAI for content, have become a key concern for Advertisers because of their inefficiency.
Points of Intervention and Limitation
As all public-facing AI providers have learned, at some point or another, intervention from the system’s developer is needed. Aligning early as an industry on where AI should be limited (by the system developer) would set up the technology for long-term viability, preempting backlash when real, provable harms occur (e.g., all driverless cars being disabled revoked after an accident). For instance children, who are less likely to understand how AI is responsible for the advertising content they’re seeing, are more susceptible to potential harm. Similarly, politics and health are two verticals where preserving the consumer’s choice to engage should be weighed more heavily.
While we are still in the early days of the industry discussion of AI in advertising, starting the conversation on principles for responsible AI use will help the longevity of the technology and protect consumers from the get-go. As a marketing practitioner, it is up to you to understand the limitations that should apply to your business to protect your brand and the precautions your partners take using AI to protect your customers. Ultimately, proactively applying principles of responsible AI in advertising will prevent both friction, when you need to implement guardrails later, and the cost (the very real impact of fines and reputational harm) of building without consideration.
Full Principle Comparison Chart
| Principle Title | AI Principle Definition | Advertising AI Principle |
| Accountability/Liability | Accountability refers to the idea that developers and deployers of AI technologies should be compliant with regulatory bodies, also meaning that such actors should be accountable for their actions and the impacts caused by their technologies. | All advertising systems that utilize AI should be compliant with the regulatory bodies in the regions they’re deployed in. |
| Beneficence/Non-Maleficence | Beneficence and non-maleficence are concepts that come from bioethics and medical ethics. In AI ethics, these principles state that human welfare (and harm aversion) should be the goal of AI-empowered technologies. Sometimes, this principle is also tied to the idea of Sustainability, stating that AI should be beneficial not only to human civilization but to our natural environment and other living creatures. | AI that is deployed within the advertising ecosystem should strive towards efficiency. |
| Children & Adolescents’ Rights | The idea that the rights of children and adolescents must be protected. AI stakeholders should safeguard, respect, and be aware of the fragilities associated with young people. | The use of AI in advertising should be limited for sensitive verticals and audiences, specifically when the target is children. |
| Dignity/Human Rights | This principle is based on the idea that all individuals deserve proper treatment and respect. In AI ethics, respect for human dignity is often tied to human rights (i.e., Universal Declaration of Human Rights). | N/A |
| Diversity/Inclusion/Pluralism/Accessibility | This set of principles advocates the idea that the development and use of AI technologies should be done in an inclusive and accessible way, respecting the different ways that the human entity may come to express itself (gender, ethnicity, race, sexual orientation, disabilities, etc.). This principle is strongly tied to another set of principles: Justice/Equity/Fairness/Non-discrimination. | All advertising systems that utilize AI ought to explicitly consider diversity and ensure that the outcomes are inclusive. |
| Freedom/Autonomy/Democratic Values/Technological Sovereignty | This set of principles advocates the idea that the autonomy of human decision-making must be preserved during human-AI interactions, whether that choice is individual, or the freedom to choose together, such as the inviolability of democratic rights and values, also being linked to technological self-sufficiency of Nations/States. | The use of AI in advertising should be limited for sensitive verticals and audiences, specifically when the topic is politics. |
| Human Formation/Education | Such principles defend the idea that human formation and education must be prioritized in our technological advances. AI technologies require a considerable level of expertise to be produced and operated, and such knowledge should be accessible to all. This principle seems to be strongly tied to Labor Rights. The vast majority of documents concerned with workers and the work-life point to the need for re-educating and re-skilling the workforce as a mitigation strategy against technological unemployment. | All advertising systems that utilize AI ought to disclose the usage to users in layman’s terms. |
| Human-Centeredness/Alignment | Such principles advocate the idea that AI systems should be centered on and aligned with human values. AI technologies should be tailored to align with our values (e.g., value-sensitive design). This principle is also used as a “catch-all” category, many times being defined as a collection of “principles that are valued by humans” (e.g., freedom, privacy, non-discrimination, etc.). | N/A |
| Intellectual Property | This principle seeks to ground the property rights over AI products and/or processes of knowledge generated by individuals, whether tangible or intangible. | All advertising systems that use AI ought to respect intellectual property rights. |
| Justice/Equity/Fairness/Non-discrimination | This set of principles upholds the idea of non-discrimination and bias mitigation (discriminatory algorithmic biases AI systems can be subject to). It defends the idea that, regardless of the different sensitive attributes that may characterize an individual, all should be treated “fairly”. | All advertising systems that use AI ought to have an option to enable any user to opt-out. |
| Labor Rights | Labor rights are legal and human rights related to the labor relations between workers and employers. In AI ethics, this principle emphasizes that workers’ rights should be preserved regardless of whether labor relations are being mediated/augmented by AI technologies or not. One of the main preoccupations pointed out when this principle is presented is the mitigation of technological unemployment (e.g., through Human Formation/Education). | N/A |
| Open source/Fair Competition/Cooperation | This set of principles advocates different means by which joint actions can be established and cultivated between AI stakeholders to achieve common goals. It also advocates for the free and open exchange of valuable AI assets (e.g., data, knowledge, patent rights, human resources) to mitigate possible AI/technology monopolies. | N/A |
| Privacy | The idea of privacy can be defined as the individual’s right to “expose oneself voluntarily, and to the extent desired, to the world.” In AI ethics, this principle upholds the right of a person to control the exposure and availability of personal information when mined as training data for AI systems. This principle is also related to concepts, such as data minimization, anonymity, informed consent, and other data protection-related concepts. | All advertising systems that use AI ought to uphold individual privacy. |
| Reliability/Safety/Security/Trustworthiness | This set of principles upholds the idea that AI technologies should be reliable, in the sense that their use can be verifiably attested as safe and robust, promoting user trust and better acceptance of AI technologies. | All advertising systems that use AI should be evaluated for safety/trustworthiness. |
| Sustainability | This principle can be understood as a form of “intergenerational justice,” where the well-being of future generations must also be counted during AI development. In AI ethics, sustainability refers to the idea that the development of AI technologies should be carried out with an awareness of their long-term implications, such as environmental costs and non-human life preservation/well-being. | AI that is deployed within the advertising ecosystem should strive towards efficiency. |
| Transparency/Explainability/Auditability | This set of principles supports the idea that the use and development of AI technologies should be transparent for all interested stakeholders. Transparency can be related to “the transparency of an organization” or “the transparency of an algorithm.” This set of principles is also related to the idea that such information should be understandable to non experts, and when necessary, subject to be audited. | All advertising systems that utilize AI ought to disclose the usage to users in layman’s terms. |
| Truthfulness | This principle upholds the idea that AI technologies must provide truthful information. It is also related to the idea that people should not be deceived when interacting with AI systems. This principle is strongly related to the mitigation of automated means of disinformation. | AI that is deployed within the advertising ecosystem should aim towards truthfulness of the products and services being advertised. |
