When using Artificial Intelligence (AI), a crucial question needs to be asked: how does it impact people? The 34th edition of Febraban Tech, held between the 25th and 27th of June, in São Paulo, had as its central theme precisely “The responsible journey in the new AI economy”.
The fair, which brought together debates across the entire economy and, mainly, the financial sector, is preparing to use new resources and advances in AI responsibly. Larissa Lima and Diego Guerreiro, both SAS customer advisors, led a talk at the event in which they discussed how the company has integrated, since its foundation, an ethical approach towards AI, in addition to showing how to incorporate practices to deal with these tools in a responsible on a daily basis. Analyzing data and using Artificial Intelligence ethically is crucial to ensuring that technologies benefit people in a fair and correct way. For SAS, the path involves responsibility, proactively identifying and mitigating adverse impacts; robustness, with reliable, scalable and high availability processing; privacy and security, respecting all data holders; and keeping the human being at the center of all processes.
This involves respecting the rights of the people involved, protecting the personal data that each person shares and dealing with them responsibly; ensure that only essential data for that specific objective is requested, invest in security tools to prevent leaks and always keep in mind that AI must serve humans, and not the other way around.
On this journey, it is essential to ensure that ethical practices keep up with the evolution of technologies. “These principles have been guaranteed by the company since its foundation. Even though, nowadays, there is more talk about it and, technically, it is possible to define what AI is, the core of our solutions has always made sure that we are prepared to work with a large volume of data. Currently, in this context of Big Data, we obviously need to be increasingly prepared. And SAS has been able to follow this evolution over time”, explains Larissa.
Guerreiro also raises another important point: the ethical issue needs to be rooted in the company’s culture: “Even though there are no established rules or laws, we need to know how to use data and create and maintain internal conduct.” Larissa adds: “companies know and understand that those who deal with AI are automatically dealing with sensitive customer data. So, there is no point in developing the best possible technology if you are not careful. It is a matter of concern for the image of the company itself, which needs to convey reliability. If the customer doesn’t trust, the company won’t evolve.”
Experts highlight that SAS works with “human beings at the center of applications”. In practice, this premise is directly linked to the ethical vision of the business. “It’s about the importance of making Artificial Intelligence develop without losing focus on ensuring that the tasks, algorithms and solutions generated by it are responsible and based mainly on respect for human rights”, defines Larissa. “It is technology benefiting human beings, not us serving it”, adds Guerreiro.
For them, one of the biggest challenges related to the use of AI today is knowing what to use, why to use it and how to use it. “It is important to question what the purpose of this use is, whether it is something that will really bring a benefit or whether the person is just getting into unnecessary hype. It’s about thinking about what the real benefit would be. We have discussed this a lot, especially using it with caution. I believe there is a cultural gap in this case. Sometimes, the ‘how to use it’ and the ‘why to use it’ end up getting lost along the way”, observes the customer advisor.
Adapting to technology
Discussions about Artificial Intelligence eliminating job positions appear at the main debate tables on the evolution of technology. On the other hand, both SAS professionals argue that people’s fears should not be that it will “steal” their jobs, but rather that they will not adapt to the correct way of using AI.
“When we say that positions will undergo changes, it is in this sense. Things are updating, modernizing, and we have to keep up, even thinking about the technical background. You need to train yourself to carry out that activity. Of course, changes will happen, but they will bring new opportunities. It’s a process”, says Larissa.
For Guerreiro, possible concerns must really involve this issue of adapting humans to the use of AI – which, despite having extremely advanced technologies, will still need the “human touch” along the way: “You need to have this human work in the end . It may be that the AI does a cool project, but the result itself is very robotic. The human being is important to make a correction, change an idea that may have been biased, make adjustments and put forward your own ideas.”
Regulatory challenge
Eventually, the conversation comes to the issue of regulation – or, rather, the lack thereof. “Another very relevant market challenge is to encourage this regulation. We want to use AI, but we hear few conversations about how we will regulate it”, points out Larissa. The creation of laws around the subject is essential, guaranteeing ethical principles, safety and responsibility, among other premises mentioned by the executives. “It’s an agenda that should always be in vogue, especially among developers, so that we can reach a common denominator”, he warns.
She also comments that Artificial Intelligence, since it is based on a database created by individuals, can end up reproducing the behaviors of these people, which opens up gaps for possibly racist or sexist analyses, for example. Hence the even more latent importance of complete and detailed regulation, which works to mitigate this type of situation, making it really only look at the customer’s consumption profile – and not at sensitive data – and which guarantees an environment free from this type of judgment. “The data involved in the analysis must only be that which interferes with the result. These are points of attention, and the best way to resolve them is to always talk about them”, he says.
In the financial market, these applications take place in processes such as credit, management and risk analysis; fraud detection; model management; prevention of money laundering; among other possibilities. Technology advances at high speed, and as a result, SAS tools become increasingly robust and complete. And precisely by having all these premises for responsible use defined and already well consolidated internally, the company is able to ensure that technology does not go beyond ethics.
( fonte: MIT Technology Review )
