Responsible AI Beyond Compliance: A Conversation on Cybersecurity, Ethics, and Technological Sovereignty with Michael Tchuindjang
We had the pleasure of speaking with Michael Tchuindjang, an Associate Lecturer at the University of the West of England (UWE Bristol) and a UK Cyber Security Council assessor. His work sits at the intersection of artificial intelligence, cybersecurity, and education, with a particular focus on strengthening the resilience of large language models against emerging security threats.
In this episode of the Teach Somebody Responsible Tech Collective podcast, we explored what responsible AI means beyond regulatory compliance, the growing cybersecurity risks associated with AI, and how governance approaches differ across Europe, the United States, and Africa. We also discussed the importance of technological sovereignty and the role of education in building a safer, more inclusive digital future.
Note: This article is based on an AI-generated Zoom transcript, edited and summarized to highlight key topics covered during the discussion.
What Does Responsible AI Really Mean?
For Michael, responsible AI goes beyond ethical guidelines and regulatory compliance. It involves anticipating how AI systems might fail, be misused, or cause unintended harm.
He argues that discussions about responsible AI often focus on whether systems meet established standards, rather than whether they can withstand real-world security threats and unexpected situations.
This is why he emphasizes security by design, where safeguards are integrated from the earliest stages of AI development rather than introduced after vulnerabilities emerge.
Michael also highlights the importance of maintaining human oversight. As AI systems become increasingly sophisticated, people may begin relying on them without questioning their outputs or understanding their limitations.
He believes AI should serve as an amplifier of human capabilities rather than a replacement for critical thinking and human judgment.
Is the Technology Industry Moving Fast Enough on AI Safety and Security?
Michael believes the technology industry is advancing much faster in innovation than in security and responsible governance.
He describes a recurring pattern in which companies rapidly release new technologies, followed by reactive security improvements when vulnerabilities or harmful consequences become apparent.
Rather than addressing risks after deployment, he argues that developers should anticipate potential threats and embed stronger protections throughout the development process.
He also recognizes AI's benefits in cybersecurity. For example, AI can assist security operations centre (SOC) analysts in processing large volumes of security data and identifying potential threats.
However, these same technologies can also be exploited by malicious actors.
For Michael, the challenge is ensuring that technological progress is accompanied by equally strong commitments to security, accountability, and user protection.
How Does AI Governance Differ Across Europe, the United States, and Africa?
Michael emphasizes that AI governance is shaped by regional priorities and socioeconomic realities.
In Europe, discussions frequently centre on privacy, digital rights, regulatory compliance, and accountability. In the United States, the focus often leans toward innovation, market competitiveness, and technological leadership.
Across African countries, however, conversations are frequently shaped by digital inclusion, affordability, cybersecurity, access to education, and infrastructure development.
These differences raise important questions about whether international regulatory frameworks can be applied equally across different regions.
Michael points to the EU AI Act and General Data Protection Regulation (GDPR) as useful examples of frameworks that establish principles around transparency, accountability, and risk management.
However, he cautions that adopting these frameworks without considering local economic conditions and institutional capacity may create additional challenges.
He advocates for governance approaches that build on international best practices while remaining adaptable to local contexts.
Why Is Technological Sovereignty Important for Africa?
One of the most important themes of our conversation was the need for greater technological sovereignty across Africa.
Michael argues that African countries should not remain dependent solely on technologies developed elsewhere, particularly when those technologies may not reflect local needs and realities.
For example, systems that require frequent internet connectivity and online updates may present challenges in communities where connectivity remains unreliable or expensive.
He suggests exploring locally developed or adapted technologies, including open-source operating systems that could function more effectively in low-connectivity environments.
For Michael, technological sovereignty is about more than developing software or infrastructure. It is about creating the freedom to innovate, adapt, and govern technology independently.
He believes that investing in local technological capabilities will be essential for strengthening Africa's digital resilience and ensuring that African countries play an active role in shaping their technological futures.
What Cybersecurity Risks Should Everyday AI Users Be Aware Of?
As tools such as ChatGPT, Gemini, and Claude become increasingly integrated into everyday life, Michael highlights several cybersecurity concerns.
One major risk involves the exposure of sensitive or confidential information when users share personal details, organizational data, or private documents with AI systems.
He also discusses adversarial attacks, particularly prompt injection and jailbreaking, where attackers attempt to manipulate AI systems into bypassing their safeguards.
Michael's research focuses on multi-turn jailbreaking, an emerging security challenge where attackers use a series of conversational interactions to attempt to circumvent an AI model's protections.
Unlike attacks involving a single malicious prompt, these techniques exploit the conversational nature of large language models.
He also highlights the growing concern around Shadow AI, where employees use unauthorized AI applications within organizations, potentially exposing sensitive information to security risks.
These challenges reinforce the importance of cybersecurity awareness, secure AI development, and responsible technology use.
Why Are AI Literacy and Cybersecurity Education Essential?
Throughout the conversation, Michael emphasizes that cybersecurity is no longer solely the responsibility of IT professionals.
As AI becomes increasingly integrated into education, business, and everyday activities, individuals need to understand both its benefits and limitations.
AI literacy involves more than knowing how to use AI tools. It also requires understanding potential risks, protecting sensitive information, questioning AI-generated outputs, and maintaining independent judgment.
Michael believes that education is particularly important for strengthening digital resilience across African countries, where differences in access to resources and infrastructure can influence how technologies are adopted.
He also recognizes the growing contributions of African cybersecurity professionals and advocates who are helping expand awareness and build local expertise.
For Michael, investing in education and practical cybersecurity skills is fundamental to creating a safer and more inclusive technological future.
What Needs to Change for a More Responsible AI Future?
Looking ahead, Michael identifies three priorities for creating more secure, inclusive, and trustworthy AI systems.
First, security by design must become a standard practice. Developers should proactively anticipate vulnerabilities and misuse rather than relying on reactive security improvements.
Second, AI governance must become more globally inclusive. Regulatory approaches should reflect diverse regional realities, ensuring that developing economies have a meaningful voice in shaping international technology policies.
Third, AI literacy and cybersecurity education must be strengthened. Individuals, organizations, and policymakers need the knowledge and skills to understand AI risks and use these technologies responsibly.
Michael also stresses the importance of international collaboration, recognizing that responsible AI requires contributions from multiple disciplines, communities, and regions.
Final Thoughts
Our conversation with Michael highlights an important message: responsible AI is not simply about developing powerful technologies, but ensuring they remain secure, inclusive, and beneficial to society.
From anticipating emerging cybersecurity threats to strengthening technological sovereignty across Africa, building a responsible digital future requires proactive security measures, context-sensitive governance, and continued investment in education.
As AI continues to transform how we live, learn, and work, human judgment and accountability must remain central to technological progress.
Ultimately, the future of responsible AI depends not only on what technology can achieve, but also on how thoughtfully we design, govern, and use it.
Join the Teach Somebody Responsible Tech Collective for more conversations exploring technology, ethics, cybersecurity, and global perspectives, with a particular focus on the Global South.