AI Training Programs for Ethical AI Use

Generative AI (GenAI) makes it easier than ever to generate text, images, videos and voice quickly and efficiently via a simple prompt. However, this advancement brings along some serious concerns around plagiarism and intellectual property (IP) infringement.

While these issues impact schools and enterprises differently, both must navigate this uncharted territory to preserve integrity and avoid legal consequences. In this blog, we will explore how AI-driven plagiarism affects the education system before transitioning into its more complex implications in the corporate world. This includes the role of AI Training in promoting ethical AI use.

I am closely involved in this topic, both as an AI lecturer at NUS and through my daughter's recent start at the University of Maastricht in the Netherlands.

AI Detection Tools in Education

Written assignments have long been the gold standard for evaluating critical thinking and subject mastery. Today, ChatGPT can draft essays, research papers, and creative stories in less than 60 seconds. This poses a significant dilemma for educators, whose primary method of assessing student understanding traditionally relies on written assignments. When students submit AI-generated work, it undermines the educational process, raising concerns about academic integrity.

Many educational institutions use AI-detection tools like GPTZero, Originality and Turnitin. I suggest to try them out yourself but none of them has impressed me. These tools use AI to estimate the likelihood that a text was generated by ChatGPT. These tools face challenges with false positives and false negatives. Improving false positives comes at the expense of false negatives and vice versa. GenAI models can mimic human writing with impressive accuracy. Moreover, students savvy enough to tweak the AI-generated content can easily evade detection, raising concerns about the efficacy of these tools.

I can confirm that many students are hesitant to use ChatGPT for assignments due to the risk of failing grades or even suspension, and this fear appears to be effective. The deterrent effect of schools potentially using these detection tools is very real!

The real solution lies in assessment models that are less susceptible to AI-driven plagiarism, like in-person assessments, oral exams, and long-term projects. These approaches require students to demonstrate their understanding in ways that AI tools struggle to replicate.

AI and Content Authenticity in Enterprises

While schools are wrestling with the challenge of evaluating original thought, businesses are grappling with similar issues, albeit on a much larger and more legally complex scale. In the corporate world, the stakes for AI-driven plagiarism and IP infringement are far higher. Companies risk potential lawsuits, brand damage, and loss of consumer trust.

The same AI tools that help students complete their homework are now being used by employees to draft reports, write marketing materials, and produce client proposals. While these tools boost productivity, they introduce risks related to plagiarism and copyright infringement. AI content generation tools are increasingly integrated into marketing automation platforms and sales automation workflows, amplifying both the benefits and risks across enterprise operations.

In contrast to the classroom, AI-driven plagiarism in the corporate world can have far-reaching effects. The stakes are especially high in industries like journalism, law, and finance, where original thought and intellectual property are crucial.

Legal Case: Getty Images vs. Stability AI

While ChatGPT specializes in text generation, GenAI has advanced into image generation through what we call diffusion models. These models transform prompts (text) into images.

Getty alleges that Stability AI used millions of its images without permission to train their AI model, thus infringing on its intellectual property rights. Getty argues that this unauthorized use constitutes an illegal reproduction and distribution of its images.

The central legal issue is whether Stability AI's use of copyrighted images for training its generative AI models falls under "fair use" or whether it violates copyright laws. Stability AI has argued that the use of these images is transformative, as the model creates new content rather than directly copying the images. The case is closely watched and is still ongoing.

Legal Case: Thomson Reuters Enterprise Centre GMBH v. ROSS Intelligence Inc.

On February 11, 2025, a Delaware federal court issued the first major decision concerning the use of copyrighted material to train AI. Thomson Reuters, the owner of Westlaw, sued Ross for using Westlaw headnotes—summaries of key points of law and case holdings—to train a competing, AI-driven legal research search engine. The court granted Thomson Reuters’s partial motion for summary judgement on its direct infringement claim and rejected Ross’s "fair use" defense.

Copyright Infringement: A Growing Risk for Enterprises

Beyond plagiarism, enterprises face the additional challenge of managing copyright infringement risks related to AI-generated content. AI models are often trained on large datasets, some of which include copyrighted material like articles, music, and artwork. If these datasets are used without proper attribution or compensation, businesses could find themselves in violation of copyright laws. It is one of the main reasons that companies block ChatGPT use.

To mitigate the risks associated with AI-driven plagiarism and copyright infringement, businesses can adopt several proactive strategies:

• AI Detection Tools: I would not recommend that companies rely on them, mainly due to the risk of false positives, which can cause unnecessary and painful disputes
• Clear Guidelines and Training: All companies should have an AI policy by now with clear guidelines for the ethical use of AI tools. Employees should be encouraged to disclose when AI has been used to assist in content creation. This transparency can help mitigate risks while maintaining trust with stakeholders.
• Licensing Agreements: Companies could explore licensing agreements with content creators to correctly use copyrighted material

As AI becomes more integrated into business operations, legislative bodies are beginning to step in. The European Union's AI Act emphasizes the importance of intellectual property protection, including copyrights. The Act requires providers of general-purpose AI models to ensure compliance with EU copyright law. It mandates that providers must put in place policies that respect copyright and provide detailed summaries about the content used for training these models, ensuring transparency while balancing the protection of trade secrets. This aims to prevent the use of copyrighted content without proper authorization during AI model training and deployment.

For example, assume that you used Midjourney to generate an image that will be used in an upcoming marketing campaign. The copyright ownership for that generated image is not automatically granted to you or the company you work for. According to Midjourney's terms of service, users generally do not own the copyright to the images they generate. Midjourney retains ownership, but they grant users a broad license to use the images as long as the use complies with their terms. If you have a paid subscription, Midjourney grants you more rights, including the ability to use the generated images commercially. However, even in this case, Midjourney still retains certain rights to the content, and you still do not fully own the generated image. For truly exclusive rights or copyright, further licensing or custom agreements may be required.

The complexity of AI-related IP risks requires organizations to develop comprehensive governance frameworks. Different jurisdictions are taking varying approaches to AI regulation and IP protection. While the EU has implemented comprehensive legislation, other countries like Singapore are developing their own frameworks, creating a complex compliance landscape for multinational organizations.

Conclusion

The rise of GenAI has brought both exciting opportunities and significant challenges in education and business. The education system is focused on preserving academic integrity while companies face more significant risks, including legal battles over IP infringement and reputational damage from AI-driven plagiarism.

Ultimately, the path forward requires balancing AI's capabilities with ethical and legal standards to unlock its full potential while protecting individual and institutional rights. Explicit labelling of AI-generated content can also help distinguish machine-made work from human creations. Strong AI Training programs are essential to ensure compliance and awareness across all organizational levels.

For all your AI questions contact Rainmakers SG!