Governments worldwide are concerned about the misuse of generative AI to create inappropriate or harmful images. Think of realistic but fake nudes or manipulated photos that can harm people. In Hong Kong, the parliament recently discussed the steps needed to curb this type of AI abuse, both legally and practically. This discussion concerns not only criminal behavior but also the responsibility of platforms, schools, and tech companies, as well as the protection of citizens and young people. This is relevant for your company: the same technology you use for process improvement, marketing, or automation can also be misused. This article explores what is at play, what the risks are, and above all: how you, as a Flemish or European SME, can deal with AI wisely, without panic and without naivety.
What exactly is going on
During the parliamentary session in question, the focus was on the use of AI to create and distribute indecent or inappropriate images. The government received questions regarding how to better protect citizens, and particularly young people and women, against these new forms of image abuse. The discussion focused on various tracks: existing legislation regarding obscenity and online crime, the role of the police and the judiciary in investigation and prosecution, and the responsibility of online platforms and service providers to detect and remove harmful content more quickly. Education and awareness were also addressed: schools and young people must learn what risks exist, how to recognize abuse, and where to turn for help. In short, work is being done on a mix of regulation, enforcement, prevention, and cooperation with the technology sector.
Impact on people and society
The misuse of AI to create indecent images directly impacts trust: trust in what you see online, but also trust in one another. For victims, the circulation of realistic fake images of them can have an enormous psychological and social impact. Even though this concerns a non-European context, the mechanisms are the same everywhere. Organizations are receiving more questions from employees, parents, students, and customers regarding digital safety and online boundaries. At the same time, this debate shows that societies are not powerless: legislation is being adapted, law enforcement agencies are learning to work with AI, and clear standards are emerging regarding what is and is not acceptable. For companies, this signals that AI is not just a matter of efficiency, but also a journey of responsibility: how do you ensure that your applications do not contribute to abuse, but rather help protect?
Ethical and sustainable considerations
Various ethical and sustainable themes converge in AI related to image generation. First, human dignity: systems capable of creating indecent or degrading images touch upon the core of respect for human beings. This gives rise to a clear duty to build boundaries into tooling and processes. Transparency is a second element: users must know what an AI system is capable of, what limitations exist, and what content is not permitted. From the perspective of sustainability and energy consumption, it is important not to simply run unlimited generative models for trivial or harmful applications: every AI call consumes energy, and behind every model stands an infrastructure with an ecological footprint. Fairness and bias also play a role: a model that more easily sexualizes or stigmatizes specific groups or bodies reinforces existing inequalities. And naturally, safety is paramount: access to sensitive functions (such as unfiltered image generation) must be controlled, logged, and restricted, especially in commercial or educational environments. An ethical AI strategy therefore means making conscious choices: what do we do, what do we explicitly not do, and how do we document this?
Safety and risk dimension
The misuse of AI for indecent imagery is one manifestation of a broader set of security risks. While generative models can create images, they can often also be used for deepfakes, phishing, social engineering, or reputational damage. Added to this are classic risks such as hacking, data leaks, and privacy violations. If your organization uses AI models or integrates them into products, you must consider: who is allowed to generate what, what data is stored, how logging and monitoring are managed, and what happens in the event of misuse? Working in a solution-oriented manner means, among other things: setting up technical filters (content filters, blocklists, moderation tools), strong access control (authentication, roles, permissions), clear rules of conduct for employees and customers, and a clear incident process if something does go wrong. By having this foundation in order, you reduce the chance that your AI environment becomes a springboard for unwanted or illegal use.
What does this mean for your business?
For Flemish and European SMEs, the most important lesson is: AI is no longer a toy, but a fully-fledged business component that requires governance. Even if your company does not offer image generation, you likely work with tools (copilots, cloud platforms, marketing software) that include these types of functions. Customers, partners, and employees also expect you to have thought about this. The European AI Act and existing privacy legislation (such as GDPR) place extra emphasis on risk management, transparency, and documentation. That sounds daunting, but it can be approached very pragmatically: start with a simple AI register (which tools do we use, for what purpose, and with which data?), define clear guidelines (what do we not do with AI?), and link this to your existing IT and privacy policies. This way, you build a mature and human-centered AI approach step by step, without big words or expensive processes.
3 concrete recommendations
- Create an AI code of conduct: Document in a maximum of 2 pages what employees are and are not allowed to do with AI (e.g., do not enter sensitive data, do not generate indecent or discriminatory content, respect for privacy).
- Choose your tools consciously: Preferably use AI platforms that offer content filters, logging, and European or equivalent privacy guarantees, and contractually stipulate what they do to prevent abuse.
- Integrate AI into your risk management: Explicitly incorporate AI into your existing information security and GDPR policies: access rights, retention periods, data minimization, and a clear incident response plan.
End
The discussion surrounding AI and indecent imagery shows that technology is never neutral: it can cause harm, but also help protect, detect, and prevent. When you, as a company, make clear choices – people over technology, with attention to ethics, safety, and sustainability – AI becomes a reliable colleague instead of a risk. At Canyon Clan, we work every day on down-to-earth, safe, and human-centered AI and software solutions that improve processes without undermining your organization's values. Would you like to explore together how you can deploy AI wisely, with an eye on regulations, ethics, and practical feasibility? Feel free to contact us for an exploratory conversation.