How to Tackle Trust, Governance, and Privacy of DAM in the Era of Generative AI

This article has been provided by Dietmar Rietsch, CEO of Pimcore.  He is a passionate entrepreneur who has been designing and realizing digital projects for more than 20 years.


It is no secret that generative AI is making inroads in almost all business activities. From creative augmentation, operational efficiency to experience improvement, brands are adopting this new revolution at an unprecedented rate. Digital asset management software is no different.

While AI can streamline and improve the digital asset creation process, there are concerns about ownership and copyright infringement. Using generative AI in DAM software includes the following: strengths of automation and efficiency, weaknesses in potential ownership and copyright infringement issues, opportunities for innovation and expanding creative options, and threats of legal disputes and the need for proper legal documentation.

Moreover, generative AI-produced work can lead to potential legal issues, including concerns about infringement, rights of use, ownership of AI-generated works, and the use of copyrighted and trademarked material without permission.

Certain lawsuits, such as Andersen v. Stability AI et al., have already been filed to clarify what constitutes a ‘derivative work’ and rely on the fair use doctrine. The outcome of these cases could shape how generative AI can explore existing works. Uncertainties in the legal landscape pose challenges for companies using generative AI, including the risk of violations and unintentional disclosure of confidential information.

Additionally, Generative AI may imitate established brands, logos, or trademarks, potentially infringing on trademark rights. Ownership of the generated digital content can also be uncertain, causing conflicts over rights and commercial use.

Tackling these challenges will require the establishment of clear guidelines and regulations to define the responsibilities and liabilities of developers and users, striking a balance between encouraging innovation and safeguarding intellectual property rights. Hence, nurturing the level of confidence that customers have in the security, privacy, and reliability of digital assets is critical, as it can lead to the following:

Loss of customer trust and loyalty – Customers may perceive the AI system and the company behind it as untrustworthy, unethical, and lacking authenticity, prompting them to seek alternatives and potentially damaging their reputation.

Exposure to legal liability for non-compliance with regulations – Using generative AI to produce content can face legal consequences and potential lawsuits for breaches of intellectual property laws.

Copyright infringement – Generated content that copies or closely resembles existing copyrighted works without permission can be considered a copyright violation. Copyright holders can take legal action against companies that infringe on their rights, leading to legal disputes and potential financial penalties.

Revenue loss – Products or content that infringes on others’ intellectual property rights may be forced to halt or modify the operations, resulting in reduced revenue. And the legal battles and damaged reputations can deter potential customers and business partners, further impacting business returns.

Brand reputation damage – Engaging in copyright breaches or being associated with the unauthorized use of intellectual property can lead to negative publicity and public perception. Consumers’ mistrust can negatively impact a brand’s reputation and public perception.


Positive side: How does Generative AI impact DAM?

Generative AI, with its innovative capabilities, brings favorable impacts to Digital Asset Management (DAM). It infuses innovation and efficiency into DAM by revolutionizing the management and utilization of digital assets. As a result, it dramatically benefits DAM through these aspects:

Automated metadata tagging: Generative AI assists in automatically tagging and categorizing digital assets, making it faster and more efficient to organize and retrieve assets within a DAM system. By analyzing the content of the assets, AI algorithms can generate accurate and relevant metadata, reducing the manual effort required for tagging.

Speedy asset augmentation and variation: Generative AI algorithms quickly create variations of existing assets. It enables DAM systems to offer a broader range of asset options without manual intervention. For example, an AI model could generate different color schemes, layouts, or styles based on a given input, allowing users to access diverse asset variations instantly.

Real-time asset generation: It enables the instantaneous creation of assets tailored to specific requirements. It can be handy when users need assets that align with current trends, events, or personalized preferences. With Generative AI, DAM systems can dynamically generate assets in real-time, providing up-to-date and customized content.

Sparking innovation and creativity: DAM systems can generate unique and original assets by leveraging AI algorithms that outwork manual creation. This infusion of AI-generated content can inspire users, spark new ideas, and facilitate the exploration of novel concepts and designs.


Risks of Generative AI in DAM

While Generative AI enhances creative processes and automates content generation, it raises concerns regarding intellectual property, ethical use, and the potential for malicious manipulation. Let’s take a look at the risks associated with generative AI in DAM:

Biased, baseless, or wrong responses: The output can be subjective, unfounded, or incorrect and can reflect the biases in the training data or generate content without facts, potentially leading to misinformation or inappropriate content.

Digital assets with malicious intent: If digital assets with malicious intent are generated using deepfakes or deceptive content, it can be harmful to individuals, organizations, or society as a whole.

Unfit for purpose with random responses: Due to the randomness inherent in the model, the generated content might not align with the desired outcome or fail to meet the specific requirements of the DAM system.

Data privacy and copyright issues: As the AI system relies on large amounts of data for training, it raises concerns regarding data privacy and copyright. If the AI model is trained on copyrighted materials or personal data without proper consent, it can lead to legal and ethical challenges.


How to create a safe way forward

Creating a safe way forward with Generative AI in the context of intellectual property issues is crucial to ensure fair use of digital assets, protect creators’ rights, and foster innovation. In addition, establishing governance, compliance, attribution, and legal frameworks is essential for responsible and sustainable development.

Establish a governance and compliance framework: Develop guidelines, policies, and procedures that outline the responsible use of Generative AI technology. This framework should include considerations for intellectual property rights and compliance with relevant laws and regulations.

Understand intellectual property rights: Educate stakeholders involved in Generative AI projects about intellectual property laws and regulations, including copyrights, trademarks, and patents. Ensure they understand what constitutes protected content and how to respect those rights.

Give due credit to the original creator: Implement mechanisms to acknowledge and attribute the original creators’ work when using Generative AI. It can be done through metadata, watermarks, or other means to recognize and honor their intellectual property.

Address legal liability: Work with legal experts to identify and mitigate potential legal risks associated with Generative AI usage. Understand the limits of liability and establish procedures for responding to claims of copyright infringement or other intellectual property disputes.


AI TRiSM strategy and tools in Generative AI development

Generative AI development is continuously progressing, necessitating an enterprise-wide strategy for AI trust, risk, and security management (AI TRiSM). By applying AI TRiSM principles to DAM systems, enterprises can enhance privacy protection, minimize the risk of data breaches, and establish trust among their users and stakeholders.

To address this, there is an immediate need for specialized AI TRiSM tools that can regulate data and process flows between users and organizations hosting generative AI foundation models. Readily available tools are necessary to ensure users have privacy assurances and effective content filtering. AI developers must collaborate with policymakers, and potentially new regulatory authorities, to establish oversight and risk management policies and practices for Generative AI.


This article has been provided by Dietmar Rietsch, CEO of Pimcore.  He is a passionate entrepreneur who has been designing and realizing digital projects for more than 20 years.

Share this Article: