The seminar organized by Piazza Copernico intends to offer tools to increase the awareness of business MANAGERS and start-up FOUNDERS of the potential risks of Criminal liability (L. 231/01), of Class Action (L. 31/2019) and violation of Privacy (GDPR 2016/679) related to the use of ARTIFICIAL INTELLIGENCES (AI) DO NOT ETHICALLY COMPLY with EU Guidelines and OECD Recommendations: anthropocentric; reliable; fair.
The event is sponsored by
The recent Guidelines of the EU ("Ethics of Artificial Intelligences"), of the OECD ("Guidelines on Artificial Intelligences") and of the European Privacy Authorities ("Recommendations for the conformity of Artificial Intelligences to the GDPR") forcefully place the need to protect companies from potential risks and / or damages deriving from the use of AI that is not ethically compliant in the provision of their services or products: criminal liability; Class action; Privacy violation; Work safety; etc..
Furthermore, the fact of being able to guarantee that the AI used will not harm man, will support the environment, will not cause discrimination (by race, sex, religion, socio-economic class, etc.), will protect user privacy (staff, trade union , health and politics), can be a decisive added value in the development of the "corporate brand" and in building a relationship of "trust" with the community of its consumers.
Therefore, it becomes necessary to increase awareness of the top management levels of the company or of the startuppers about the strategic value of acquiring (ex-ante; in itinere; ex-post) of CONTROL TOOLS AND ETHICAL VERIFICATION OF AI developed or acquired in the company.
Each project must be able to guarantee a significant level of:
The primary purpose, therefore, is to prevent:
• Financial losses due to self-referencing automatic trading loops;
• Errors in industrial investments due to incorrect big data (bias);
• Mass electronic identity theft or systematic violations of privacy;
• racial or social or religious discrimination in the provision of credits or loans;
• Gender discrimination in the evaluation of curricula or in personnel management;
• Fatal accidents at work in Industry 4.0 production cycles;
• fatal accidents deriving from autonomous driving systems (land and air);
• Fake news and hate campaigns on social networks or telematic information platforms;
Adopting an ethical model therefore becomes a strategic choice that guarantees the global competitiveness of the product / service that uses AI, because:
a) reassures the user communities that will use it;
b) lowers the risk of commercial boycott or hate campaigns on social media;
c) reduces any damages or Class Action that could derive from the damage to human beings (at single level or of entire communities̀);
d) ensures regulatory compliance;
e) increases the "reputation" of the corporate Brand.
The subjects to whom the seminar is addressed:
• COMPANY MANAGERS or START-UP FOUNDERS, committed to managing the development of AI or their use in support of products / processes / services.
• MINISTERIAL DIRECTORS or PUBLIC BODIES, which are asked to develop functions or services capable of providing citizens with a different (non-bureaucratic) public administration, capable of not discriminating against the weaker groups and in any case not causing damage / privileges to economic or entrepreneurial categories.
• BANKS and INSURANCE MANAGERS AND SPECIALISTS, in order to assess the risk of the potential policyholder or of the potential loan to a company that uses artificial intelligence for its customer services.
• RESEARCHERS or PROGRAMMERS, committed (individually or in teams) to develop innovative solutions and techniques in the area of neural networks or machine learning, so that they can evaluate the potential consequences of research results (or solutions) before they are placed on the market.
• STAKEHOLDERS, focused on identifying new areas of responsibility so far not explored, or on the function of checking and verifying the requests of their reference subjects (from trade associations to those of consumers or minorities).
• The state of the art of artificial intelligence and the potential ethical risks to be evaluated
o The different types of AI:
o The different approaches for implementing the algorithms:
- Renforced Learning,
- Neural and Generative Adversarial Networks;
o The ethical and legal aspects of an AI:
- The human-centric approach in planning, management and monitoring,
- The pillar of the guarantee is a reliable and robust AI,
- The ethical aims of AI,
- The ethics of AI as a factor of global competitiveness of the EU,
- The relationship between AI Ethics and GDPR,
- The node of responsibility and / or legal personality of AI and robots,
- The profiles of legal responsibility for companies (current legislation).
• The EU Guidelines and the OECD Guidelines on AI Ethics
o The EU coordinated plan on AI
o Reliability requirements in the implementation phase:
- Data management,
- Design for everyone,
- Management of AI autonomy and human control,
- Respect for human autonomy (decision-making) and Privacy,
- Strength, safety, transparency,
- Involvement of stakeholders.
• The AI Etichal Model Canvas
o Introduction to the Business Model Canvas
o The innovative Canvas for AI Ethics:
- The fundamentals:
• Human autonomy,
• Data governance,
• GDPR and Privacy,
• Social and environmental impact,
• Human impact,
• Wellness and benefits;
- The transversals:
• Fairness and equality,
• Robustness and security,
• Transparency and traceability;
o The use of the "Legal Design" methodology for transparency
Arch. DANIELE VERDESCA
Information architect, he is currently Director of the Building Fund of the province of Lecce. Former contract professor at the Department of Political Economy of the University of Siena and General Director of Formedil in Rome. For some time at the forefront of the promotion of the digitalization of management systems in the construction sector, he is currently involved in various innovative projects aimed at the use of artificial intelligence for automated checks and controls of the regularity of companies. Author of numerous books and technical articles on the subject.