Blog by Evelyne Para, SI Representative to UNESCO, Paris.

Today, the adoption rate of AI far outpaces that of personal computers and the Internet, with generative AI models rapidly transforming the information ecosystem.

However, many voices are warning of the risks AI poses to the integrity of information and scientifically established facts, as well as the serious threats it presents to democratic institutions and public trust more broadly. Without effective governance, the genuine opportunities offered by AI risk remaining untapped or being exploited unevenly, thereby exacerbating existing inequalities.

While there is ample evidence of AI’s propensity for bias—often stemming from incomplete or unrepresentative datasets—there are also significant concerns regarding gender and racial stereotyping, as well as serious threats to the development of critical thinking and respect for a plurality of opinions.

What can we do to address the divides and risks associated with AI that might limit its positive contribution to scientific progress, economic growth, and the achievement of the Sustainable Development Goals (SDGs)?

Work conducted by UNESCO in Paris over the past several months—in which I had the privilege of participating as an SI representative—focused on the need for state governance of digital platforms.

Here is a brief summary of this work at UNESCO.

Monitoring the governance of digital platforms and user empowerment

Digital platforms have provided individuals and societies with immense opportunities for communication, engagement, and learning. However, the potential of these platforms has gradually eroded in recent years due to a lack of foresight in managing key risks. Work initiated by UNESCO in 2023 on Principles for the Governance of Digital Platforms was closely followed during meetings held at UNESCO Headquarters in Paris in June 2026.

Testimonies from scientists and user experiences were gathered to assess how to better address the needs of populations requiring special attention—such as vulnerable and marginalised communities, women and girls, artists, and human rights and environmental defenders. Some speakers emphasised the imperative need to ground digital platform governance in human rights in order to further foster cultural expression and content diversity.

Reference was made to the work already undertaken at UNESCO—including insights from my January 2025 blog post on the Soroptimist International website, some of which was presented in Geneva in July 2026—as well as to UNESCO tools and guides for teachers and students, all aimed at preserving our educational autonomy. The same applied to the methods for evaluating the implementation of the global UNESCO Recommendation on the Ethics of AI.

The meetings I have had the privilege of attending at UNESCO over the past several months enabled me to exchange views with representatives from dozens of countries. These nations have developed strategies, monitoring processes, plans, and regulations to ensure safer and more beneficial AI use for their populations—focusing particularly on transparency, accountability, and aligning existing systems and processes with international human rights standards. These valuable exchanges also provided comparative insights, allowing countries to learn from one another.

Recommendations have been issued to technology companies and digital platform providers

In June 2026 at UNESCO, speakers emphasised the urgent need to improve information, particularly regarding geographic, cultural, and linguistic indicators. There is also an urgent need to ensure that the very architecture of the digital space stops favouring divisive content. The UN body is also calling on governments to implement greater regulation to establish “ethical artificial intelligence”. To shape the technological development of AI, we need effective governance frameworks underpinned by the ethical and moral values we all cherish.

Follow-up on the Principles—developed since 2023 through a global multi-stakeholder consultation that gathered over 10,000 comments from 134 countries—must continue to foster inclusive participation and cultivate a trusted Internet that ensures a diversity of voices is taken into account.

Benchmark industry standards must also be developed for the generative AI supply chain, including specific requirements for investors, developers, providers, and deployment companies. We need public and private sector developers who are committed to prioritizing people—rather than profits or geopolitical considerations.

In conclusion,

Throughout these meetings, participants reiterated their commitment to implementing and improving policies and regulatory frameworks to ensure that AI is used safely and responsibly, developed and applied globally in an ethical manner for the benefit of humanity and our planet, and to ensure that the development of artificial intelligence benefits everyone.