According to the latest 2025 Global Gender Gap report from the World Economic Forum, we are 123 years away from achieving gender equality. But the risk is that this deadline will move significantly further away, rather than closer, if the world of AI continues not only to reproduce gender stereotypes, but to multiply them, particularly with generative AI. This says a lot about how strategic it is to have mixed AI working groups, in which gender differences are represented, together with other differences, such as those related to ethnicity.
However, with respect to gender, we have a gap to fill, namely the female presence in STEM disciplines, and in particular in ICT, where the percentage of female graduates is stuck at around 16%. Widespread awareness-raising on these issues can make the difference in an approach that is not just defensive or alarmist, but uses the opportunities offered by AI to make stereotypes visible and counter them.

For the first time in the history of AI, more than 100 international scientists led by Turing Prize winner Yoshua Bengio have produced a global report on the emerging risks of artificial intelligence. This document represents the first systematic attempt to map the challenges that await us: from algorithmic biases to mass disinformation, from loss of control over autonomous systems to geopolitical implications. The intervention will not limit itself to illustrating the identified risks, but will explore the concrete solutions proposed by the scientific community, including computational governance and international agreements to regulate the development of AI, considering the complexities of geopolitical dynamics and economic interests. An opportunity to understand where we are today and what tools we have to guide AI towards a safer and fairer future.

Risks related to AI systems have become a recurring topic in public and academic debates, and the European proposal for the AI ​​Act explicitly adopts a risk-based, tiered approach that associates different levels of regulation with different levels of risk. However, a comprehensive framework for conceptualizing AI-related risks is still lacking. In this presentation, I will focus on the epistemological analysis of AI risks, showing how a multicomponent risk analysis, which distinguishes between the dimensions of danger, exposure and vulnerability, allows us to better understand the sources of AI risks and intervene effectively to mitigate them.

When we think about war, which particularly in this period presents us with terrible testimonies, we have in mind a traditional war model, however technologically advanced the weapons and intelligence tools (including AI) used may be. However, for some time now, what is commonly called “hybrid war” has been increasingly spreading, the cornerstones of which are propaganda and disinformation on the Internet, especially by authoritarian regimes, with clear objectives of manipulating public opinion. As users of information, it is necessary on the one hand to support ethical journalism practices (including “participatory journalism” which uses Open Source INTelligence (OSINT) techniques and on the other to understand how to select and diversify sources.

Which companies in Italy offer products/services that use digital in a responsible/sustainable way? We will find out by talking about the Responsible Digital Consumption Observatory, born from the collaboration between the Consumer Movement and Sloweb. The speech will delve into the principles underlying this initiative, will present some examples of organizations that have become part of it and how to join.

The intervention will focus on a specific experiment regarding generative AI. Preliminary reading of the material is recommended, as the intervention will focus on specific questions and not on the description of the experiment.

Artificial Intelligence is now the global reference for information. According to some sources it will be used by 700 million people by 2030, but today it is estimated that 77% of devices in use host some form of AI. Services are provided thanks to computing architectures concentrated in server farms that use enormous linguistic models. Model training and inference activities present severe ecological challenges and social and cultural risks. The dubious quality of the training questions cast doubt on the factuality of the answers. In contrast to all this, open source language models and new hardware that enhance personal and peripheral computing are becoming increasingly important. Linguistic models built on a voluntary basis also allow for agility, cultural richness and specialization. RAG technology significantly improves factuality and allows innovative models of use, for example in teaching

140 consortium entities with which it operates as in-house providers: CSI Piemonte is a private entity under public control, operating in the technological innovation sector at the service of public entities for digital transformation.
For this reason, ethics and responsibility must guide all decisions, including those aimed at achieving the best possible balance between the three ESG (Environmental, Social, Governance) dimensions of sustainability.

Digital transformation generates value only if driven by leadership that combines vision and discipline. Three pillars: 1) Data governance: quality, security, privacy-by-design, human control over algorithms, bias auditing, traceability of decisions; 2) Transparency and accountability: clear reporting of ESG results, measurable indicators, consistency between words and deeds, continuous dialogue with stakeholders; 3) ESG integration into processes: objectives in budgets and KPIs, management incentives, responsible supply chains, innovation oriented towards social and environmental impacts. Green IT as a concrete lever: efficient data centers powered by renewables, optimized cloud, “light” software and sites, eco-design of devices, smart working and IoT to reduce consumption. Call to action: leaders are ethical guardians of digital, transforming compliance into competitive advantage and lasting trust.

An in-depth look at how green technologies – from the sustainable cloud to the reduction of consumption in data centers – can generate not only environmental benefits, but also a strategic positioning on the market.

Ghost working represents a modern form of labor exploitation in digital capitalism, where millions of invisible workers perform tasks essential to the functioning of artificial intelligence systems. The phenomenon, theorized by Mary Gray and Siddharth Suri, reveals how behind the apparent full automation of AI lies a massive use of precarious and underpaid human labor. Ghost workers perform key tasks such as data labeling, content moderation, data transcription and verification, working across fragmented micro-tasks distributed across digital platforms. The intervention highlights a global geography of exploitation with enormous wage disparities between Global North and South, precarious working conditions characterized by the absence of formal contracts, social protections and constant algorithmic control.

1. Operational approach to Green IT: identification of the main levers available to reduce IT’s carbon footprint
2. Experiences in Italy and trends: evaluation of the situation in Italy with implementation cases and comparison with current trends in other countries
3. Green IT & IT4Green: complementarity of the two approaches in addressing the positive and negative externalities of IT

The speech will address the progress made by the Italian industry in adopting more sustainable approaches, analyzing in particular the challenges posed by the “Double Transition” and the different speeds between the Member States of the European Union. Obstacles will be explored, such as European hyper-regulation, compliance difficulties for Italian SMEs compared to large global players (i.e. non-proportional horizontal obligations), and the need to overcome individual national strategies that lack a unitary European vision. The objective is to push Europe to identify its own niche of sustainable development, based on reliability and reduced consumption, rather than chasing the great global powers with projects that Europe cannot maintain.