Responsible artificial intelligence depends not only on data and algorithms, but also on the people who design, oversee and scrutinise the systems. Constructive dissent, a diversity of perspectives and robust governance are key to anticipating risks, reducing bias and building ethical, safe and people-centred AI solutions.
Contact Centre Platform Architecture and Operation area
Why artificial intelligence needs more than just data and algorithms
The answer does not lie solely in algorithms or in the quality of the data. It also depends on the people who design, oversee and, above all, question these systems.
In this context, constructive dissent emerges as a key element in answering that question. Far from being an obstacle to innovation, the ability to ask uncomfortable questions, challenge assumptions and weigh up different perspectives is one of the best tools for identifying risks before they become problems.
The value of disagreement: how dissent reduces risks in artificial intelligence
For decades, influenced by increasingly short-termist views, we have associated rapid consensus with efficiency. However, when working with systems capable of affecting millions of people, premature consensus can become a
The history of innovation shows that the most costly mistakes tend to occur when no one questions the prevailing decisions. In artificial intelligence, this can result in unnoticed biases, privacy issues, a lack of transparency or unexpected consequences arising from the use of certain models.
Constructive dissent is precisely about creating spaces where divergent opinions can be expressed, based on arguments, evidence and technical analysis. It is not about creating conflict, but about enriching the decision-making process.
When a team encourages rigorous debate, it increases the likelihood of identifying blind spots, evaluating alternative scenarios and anticipating risks that might go unnoticed in overly homogeneous environments.
Transparent AI: why technology should not replace human interaction
A key distinguishing feature of Telefónica’s approach is that it does not seek to create a false impression of human interaction. The aim is not to replace people or blur the line between humans and machines, but to harness the capabilities of AI to improve the experience of customers, employees and users in a transparent manner.
When an interaction is being handled by an AI system, it must be made clear that it is a technological tool and not a human interlocutor. This transparency is essential for building trust, setting appropriate expectations and ensuring the ethical use of technology.
AI provides speed, personalisation and analytical capability, whilst people continue to contribute judgement, empathy, creativity and accountability. The best experience arises when humans and artificial intelligence collaborate in a complementary manner.
Diversity of perspectives: key to reducing bias in AI systems
One of the greatest challenges of artificial intelligence is the emergence of bias.
For this reason, in organisations that aspire to build ethical AI, such as Telefónica, development goes beyond technology and requires multidisciplinary teams capable of contributing different perspectives.
The diversity of experiences, knowledge and approaches allows us to challenge assumptions that might be taken for granted within a homogeneous group. An engineer may spot a technical limitation; a privacy expert may identify regulatory risks; a user experience specialist may anticipate social impacts; and a business manager may assess operational consequences.
Combining these perspectives leads to better decisions and helps build more robust and balanced systems.
The importance of governance in the development of responsible artificial intelligence
In large-scale technology organisations such as Telefónica, responsible AI is achieved by combining innovation, human oversight and robust governance frameworks.
This governance is not merely about defining rules or procedures. It also involves establishing mechanisms to ensure that key decisions are subject to critical review.
- Is there a risk of discrimination?
- Is the data used representative?
- Can the result generated by the model be explained?
- What happens if the response fails?
- How is people’s privacy protected?
must be a regular part of the development process.
In this context, constructive dissent acts as an additional tool at every level, helping to identify vulnerabilities before systems go live.
Critical thinking: the key skill in the age of automation
The growing capabilities of AI models might lead one to believe that the role of people will become increasingly less relevant. However, the experience of leading organisations demonstrates the opposite: value arises when people and technology collaborate.
The more advanced the tools, the more important the human ability to interpret results, question conclusions and exercise judgement.
AI excels not in ‘being’, but in ‘doing’. That is why skills such as creativity, empathy, ethical judgement, negotiation and strategic thinking remain uniquely human.
Fostering cultures that develop these capabilities is not just a cultural issue, but a strategic one. Organisations that promote professionals capable of analysing evidence and challenging hypotheses will be better prepared to harness AI without losing control.
In companies such as Telefónica, artificial intelligence does not replace talent, but rather amplifies it.
Leadership in the age of AI: trust, debate and accountability
The adoption of responsible AI also transforms leadership.
Leaders must go beyond simply driving results: they must create environments where people feel confident about both the outcome and the process.
Expressing well-founded disagreements, raising risks and questioning decisions becomes a strategic asset. When opinions are valued for their technical rigour, the quality of decisions improves and a culture of accountability is strengthened.
This approach allows for a balance between rapid innovation and high ethical standards.
Conclusion: the best artificial intelligence is the kind that can be questioned
Responsible artificial intelligence does not arise solely from better algorithms, but from better conversations.
Organisations that promote debate, a diversity of perspectives and critical thinking are better equipped to build secure, transparent and trustworthy systems.
In an environment where technology is advancing rapidly, the real differentiator is not just developing AI, but knowing how to question it when necessary.
Because the most valuable innovation does not eliminate disagreement, but rather turns it into a tool for making better decisions.
Telefónica: leading responsible innovation as a gateway to digital technology
Telefónica is taking on the challenge of becoming the best gateway for citizens to access digital technologies, driving the company to be more innovative, competitive and customer-focused. This approach is underpinned by European leadership, a contribution to technological sovereignty, the development of advanced services and the construction of infrastructure that enables access to the most innovative technology.
The objective is clear: to offer more and better services, to build trust in the use of artificial intelligence and to ensure a positive impact on society, where quality is measured by people’s actual experience.
- Artificial Intelligence
- Creators
Blog16/07/2026
Artificial Intelligence
How Telefónica links AI research to customers’ real-life experiences
Blog15/07/2026
Artificial Intelligence
AI is redefining how we build products: the new limit is no longer technology – it’s us
Blog13/07/2026
Artificial Intelligence
AI doesn’t turn into a service on its own: how partnerships drive its actual delivery to the customer
Contact our communication department or requests additional material.
