Interpol observes a clear shift: cybercriminals are increasingly using artificial intelligence to make their attacks smarter, faster, and harder to detect. Examples include perfectly written phishing emails in multiple languages, credible deepfake videos of officials or business leaders, and automated attacks that adapt at lightning speed. At the same time, international police forces are building their own AI systems to detect and combat these threats.
The arms race in cyberspace sounds far removed from your SME, but in practice, it touches upon very concrete questions: how do you still recognize fake messages? How do you deal with deepfakes? And how do you ensure that AI applications in your organization remain secure? In this blog, we take a sober look at what is at play, which risks are real, and how you, as a Flemish or European company, can increase your digital resilience step by step.
What exactly is going on
Interpol reports that cybercriminals are increasingly actively deploying AI to professionalize their attacks. With the help of generative AI, they can:
- writing phishing emails without spelling or grammatical errors, tailored to the language and tone of the target;
- creating deepfake audio and video of public figures or executives, to mislead people;
- Organizing large-scale scams more efficiently, with a higher success rate.
Behind the scenes, teams at Interpol are working on countermeasures. Among other things, they are developing AI-driven systems to detect suspicious patterns, recognize fraudulent content, and intervene more quickly in cross-border cyberattacks. The message from this work is twofold: AI is a powerful tool for malicious actors, but also an essential instrument for defense.
Important: this is not science fiction, but ongoing developments that are already visible today in police work and cybercrime investigations.
Impact on people and society
The use of AI by cybercriminals has a direct impact on trust. Whereas in the past phishing could often be recognized by poor language use, messages are now polished, personal, and persuasive. Employees must therefore not only be vigilant but also learn to look at email, chat, and video in a new way.
For organizations, the nature of digital security is changing. Security is no longer just an IT issue, but also a human and cultural one: how do your people handle unexpected requests, payments, password resets, or seemingly urgent questions from a “manager” via email or video? The distinction between real and fake is becoming more subtle.
On a societal level, this puts pressure on our information ecosystems. Deepfakes and AI-generated disinformation can undermine trust in the media, government, and businesses. At the same time, Interpol's work shows that cooperation and technology are also becoming stronger on the defensive side. That balance is crucial: not a doomsday scenario, but a continuing mandate to build resilience and digital literacy.
Ethical and sustainable considerations
AI in cybercrime and cyber defense touches directly on a number of ethical and sustainable themes:
- Ethics & honesty: Deception via deepfakes, identity fraud, and the manipulation of emotions are clear violations of basic norms. On the defensive side, police services and companies alike must ensure that their AI systems are deployed proportionately, with respect for rights and freedoms.
- Transparency: How clear are you with employees and customers about where you deploy AI (for example, in fraud detection or monitoring)? Transparency helps build trust and prevent misunderstandings.
- Bias: AI systems that detect fraud patterns must not unfairly mark certain groups as suspicious more frequently. This requires attention to data quality, fairness, and regular audits.
- Sustainability & energy consumption: Both offensive and defensive AI consume computing power and energy. The question, therefore, is how to deploy AI so that the societal added value (less fraud, more security) outweighs the ecological footprint? For companies, this means choosing efficient models, making conscious cloud choices, and avoiding unnecessary AI bells and whistles.
- Security & privacy: AI security solutions often process sensitive data. Ethical use means: minimal data collection, strong encryption, clear retention periods, and respect for GDPR.
A human-centered, sustainable AI strategy therefore means deploying AI where it adds value, with regard for the environment, fairness, and the rights of individuals.
Safety and risk dimension
The risks surrounding AI and cybercrime are concrete, but manageable if you clearly identify them:
- Hacking and advanced attacks: AI can refine attacks, for example by automatically scanning vulnerabilities or guessing passwords. This increases the need for up-to-date systems, strong authentication, and good patch management.
- Data leaks: More digital processes mean more data. With weak security, a single phishing email can lead to a large-scale data breach. AI-supported attacks can be more targeted, for example, targeting financially or medically sensitive information.
- Privacy: When AI is deployed to monitor traffic or analyze behavior, there is a risk of going further than necessary. Privacy by design remains an absolute prerequisite.
- Abuse of AI tools: Publicly available AI models can be used by employees unknowingly to process sensitive business information (for example, by pasting confidential text into a chatbot). This creates new channels for leakage.
The key is a sensible, layered approach: technology (firewalls, endpoint security, logging), processes (incident response, access control), and people (awareness, clear agreements). Not by saying 'no' to everything, but by deploying AI in a controlled and thoughtful manner.
What does this mean for your business?
For Flemish and European SMEs, this evolution primarily means that digital security is no longer a nice-to-have, but a basic prerequisite for healthy business. AI is changing that landscape, but need not be a cause for panic.
Some sober observations:
- You are already confronted with AI-assisted phishing today, even if you do not explicitly notice it.
- Your employees are the first line of defense – and at the same time the most attractive target.
- Your own AI and software projects can be set up securely and human-centered, with built-in security and clear governance.
That doesn't require a massive budget or an in-house security department, but rather conscious choices: clear policies, basic IT hygiene, and attention to security, privacy, and ethics from day one in new digital projects. This is how you turn AI into an ally in process improvement, not an additional source of risk.
3 concrete recommendations for SMEs
- 1. Organize a no-nonsense security & AI awareness session
Schedule a short, practical annual session for all employees on phishing, deepfakes, and safely using AI tools. Show examples, practice with borderline cases, and agree on simple rules (for example: never make payments based solely on an email or video). - 2. Establish basic agreements regarding the use of AI
Write down in plain language which AI tools employees are and are not allowed to use, which data must never be shared externally, and how they can report any doubts. Combine this with technical measures such as two-factor authentication, role-based access, and logical permissions per job function. - 3. Build security into every digital step
Are you starting a new software or AI project? Include security, privacy, and sustainability in the requirements: data minimization, logging, access control, energy-efficient architecture. Work with partners who bring these themes to the table, rather than trying to tack them on afterwards.
Conclusion: technology at the service of people
Stories from the Interpol world show how sharp the edges of AI can be: the same technology that makes a phishing email perfect can also help protect millions of people against fraud. The difference lies in the choices we make, and in how consciously we build.
When your company focuses on clear processes, people-centric agreements, and responsible technology, AI becomes a security enhancer rather than a threat. At Canyon Clan, we help SMEs develop AI and software solutions that improve processes and are inherently secure, ethical, and sustainable.
Would you like to explore how to wisely deploy AI in your organization – with an eye for security, privacy, and humanity? Contact us without obligation. We are happy to think along with you in a down-to-earth and concrete way.