Ethical governance of artificial intelligence (AI) systems and tools is necessary. If the dizzying and disorderly growth of this technology continues, polarization will worsen, fundamental freedoms will be lost, and prejudice and discrimination will persist.
Led by UNESCO, the world's most comprehensive international framework has been developed to shape the development and use of artificial intelligence technologies. This is the recommendation on the ethics of artificial intelligence, adopted by 193 Member States in 2021. That is, we have an international consensus that is recognized and adopted by most countries in the world. We must now effectively implement the parts of the Recommendation that call on governments to create institutional and legal frameworks to ensure that these instruments contribute to the common good.
More information
This is not only about warning about the dangers of these technological developments without control and governance, but also about demanding that the immense opportunities they offer be used to address some of humanity's most pressing problems, such as: solve gender inequality.
The Commission on the Status of Women (CSW67), the highest body assessing progress in gender equality and women's empowerment at the global level, recognized in its recent meeting “the fundamental role of technology and innovation in accelerating the Gender Equality”. There was a plan for governments, the private sector, civil society, youth and children to promote the participation and equal leadership of women and girls in the design, transformation and integration of digital technologies and innovation processes.
In particular, the plan calls for the inclusion of a gender perspective in the design of new technologies and the adoption of regulations that ensure the fight against new risks, gender stereotypes and data protection violations; and improve transparency and accountability. One way to achieve this is to develop digital tools and services that respond to the needs of women and girls, particularly with regard to their education, health, economic empowerment and participation in public life, and to ensure that they are a Having access to digital literacy throughout life, as well as promoting policies to achieve gender parity in emerging scientific and technological fields and to create women-friendly workplaces and educational environments.
Achieving this requires specific laws and policies that address challenges such as: B. training those who develop these technologies and those of us who use them, digital content and awareness campaigns that challenge gender stereotypes and negative social norms and include education and empowerment of young people and children so that they become actors of gender equality.
We can also operate from other areas. Many artificial intelligence models (LLMs in English) are open source, compiled by programming communities and not owned by companies, and are therefore adaptable to organizations of all sizes and trainable with inputs according to the needs and philosophy of each organization. So there is a lot that can be done by technology companies on the one hand, and governments, civil society and businesses on the other, to train artificial intelligence tools by feeding the algorithms to produce results without sexist and racist biases. or classist, as is currently the case.
We also have the option of prompts, i.e. instructions that we give in chats with artificial intelligence. If we give them female roles, gender perspectives and unbiased and unbiased instructions, we will also achieve better quality results that are more similar to the world we want to build.
To move forward in this direction, we need three things: budget, political will and more women in the areas where these instruments are developed, but also where policies are debated and the norms that govern them are decided.
As is often the case in crises and times of change, we are faced with an enormous risk, but also a great opportunity. Given the direction this industry is taking, LLMs increasingly require high-quality information that “educates” them, representative of the diverse world we live in, especially with the proactive intention of not reproducing stereotypes or prejudices.
It is important that those of us who advocate for the rights of women and girls are there to provide this information, but also to ensure that these tools contribute to their equality and empowerment. Only in this way can we have unbiased artificial intelligence in the service of gender equality and building a just, sustainable and supportive world.
Maria Noel Vaeza is UN Women’s regional director for the Americas and the Caribbean.
You can follow Future planet In XFacebook, Instagram and TikTok and subscribe to our newsletter here.