Media: 联合早报 聯合早報 Lianhe Zaobao
Generative artificial intelligence (AI) is becoming widespread at an unprecedented pace. While businesses enjoy significant gains in efficiency, they are also facing a new challenge: as laws and regulations struggle to keep up with technological advances, how should innovation and responsibility be balanced?
In recent years, many multinational corporations have updated their AI usage policies, covering areas such as content disclosure, archival management, prompt documentation, tool approval, and procurement standards. While these may appear to be internal corporate guidelines, they actually reflect how businesses have become the de facto rule-makers of the AI era.
- Transparency and Disclosure: The New Baseline of the Right to Know
Many companies now require clear labeling of AI-generated images, videos, and audio content.
For marketers, this presents a dilemma. Some fear that disclosing AI involvement may diminish emotional impact. Yet in an era where reality and fabrication are increasingly difficult to distinguish, transparency may become a brand’s most valuable asset. In highly regulated industries such as pharmaceuticals, finance, and professional services, proactive disclosure not only reduces legal risks but also helps build long-term trust.
This approach is consistent with global regulatory trends. Discussions in the European Union, the United States, South Korea, and Vietnam have largely focused on transparency, authenticity, and accountability. By establishing disclosure frameworks early, companies are not only preparing for future regulations but also managing future compliance costs.
For the public, the right to know is equally important.
2. Archival Systems: Leaving a Trace for Every AI-Generated Output
Beyond disclosure, many organizations require all AI-generated materials to be archived in designated management systems, creating a single source of truth for tracking AI usage.
Although this may appear to add administrative burden, it is in fact a critical foundation for brand governance and risk management. When multiple teams and markets simultaneously use AI tools, the absence of centralized oversight can lead to version inconsistencies, conflicting messages, and even regulatory violations.
More importantly, when regulators raise questions, companies must be able to explain the origin of content, the tools used, and the approval process involved. Without complete records, even the most comprehensive compliance statements may become meaningless. A “single source of truth” is, in essence, a key line of defense against accountability risks.
3. The Chain of Responsibility Behind Every Prompt
Many companies require standardized records for every AI-generated asset, including the platform used, prompts, reference images, and final outputs. These records must often be reviewed and acknowledged by both content creators and business owners.
This requirement addresses one of the most controversial issues surrounding generative AI: intellectual property rights and accountability. Prompts are not merely creative instructions; they form a critical part of the content generation process. Should a copyright dispute arise, detailed records may become crucial evidence.
At the same time, requiring sign-off from business stakeholders signals that AI governance is no longer solely the responsibility of IT departments. Legal, privacy, intellectual property, and corporate communications teams must all be involved. Effective governance does not shift responsibility to technology; it ensures that accountability is clearly assigned at every stage of decision-making.
4. The Business Logic Behind Whitelists
Some companies approve only selected platforms such as Adobe Firefly, Getty Images GenAI, Shutterstock AI, and Microsoft Copilot, while excluding other popular tools. The reason is often not functionality, but risk.
Large enterprises evaluate more than usability. They consider whether training data sources are transparent, whether commercial licensing rights are comprehensive, and whether the company can mount a reasonable legal defense in the event of disputes. If the origins of training data remain contested, the potential legal costs can be difficult to quantify.
This reflects a form of corporate discipline and restraint: just because a tool can be used does not mean it should be used. Market popularity does not automatically translate into business viability. What companies truly care about is whether a technology can withstand regulatory and legal scrutiny.
5. Approval and Review Mechanisms: Balancing Innovation and Risk
For emerging tools that have not yet been approved, many companies implement formal application and review processes. Employees seeking to use platforms outside the approved whitelist must first complete risk assessments and obtain the necessary approvals.
This represents the delicate balance between innovation and control. Completely banning new technologies may mean missing valuable opportunities, while unrestricted adoption can create significant legal exposure. Through structured review processes, companies preserve room for innovation while minimizing potential risks.
In areas where regulations remain underdeveloped, the cost of “adopt first, fix later” is often far greater than that of “review first, then adopt.” In the future, an organization’s ability to assess and manage AI-related risks may itself become a competitive advantage.
Conclusion
While governments around the world continue to catch up with the rapid pace of technological development, multinational corporations have already begun building their own AI governance frameworks. From disclosure and record-keeping to review processes and accountability mechanisms, these systems ultimately serve a single purpose: ensuring that innovation and responsibility advance together.
The true test of the AI era has never been how advanced the technology becomes, but whether society can preserve trust and order. The measure of a mature society is not how intelligent its machines are, but whether it remembers a fundamental principle: machines exist to serve people, not the other way around.
Talk to us to solve your digital challenges.
#ArtificialIntelligence #AIGovernance #ResponsibleAI #AIEthics #CorporateGovernance #DigitalTransformation #InnovationManagement #RiskManagement #Compliance #TrustInAI #AIGuidelines #AgenticAI #AIWorkflow #Tech
