The future of artificial intelligence does not depend solely on the boundaries of technology, but on how we intelligently lead its unlimited potential. On May 14, 2024, the Shanghai Artificial Intelligence Industry Association (hereinafter referred to as SAIA) was invited by Ms. Hannah Teddy, Global Secretary General of the TIC International Inspection and Certification Council, and Xu Min, Chief Representative and Secretary General of the TIC International Inspection and Certification Council in China, to attend the 2024 TIC Summit held in Brussels, Belgium. At the summit, Mr. Xu Qi, Deputy Secretary General of SAIA, discussed the hot topic of "What does it take to build trust in artificial intelligence?" on behalf of SAIA with top scholars, industrial leaders and regulators in the field of artificial intelligence around the world.
The summit held a special video link with London Mayor Michael Mainelli, and offline invited Dr.-Ing. Michael Fübi, Chairman of the TIC Board of Directors, Jeremy Rollison, Senior Director of EU Policy and European Government Affairs, Alexander Hoefmans, Deputy Director of Digitalization, Privacy and Administrative Simplification at the Cabinet Office of the Secretary of State of Belgium, Alyssa Lefaivre Škopac, Head of Global Partnerships and Growth at the Responsible AI Institute (RAI), Xu Qi, Deputy Secretary-General of the Shanghai Artificial Intelligence Industry Association, Risto Uuk, Head of Research at the Future of Life Institute, Frederico Oliveira da Silva, Senior Legal Officer of the European Consumer Organization BEUC, Susan Taylor Martin, CEO of the British Standards Institution, Alexander hoequi, CEO of TUV Rheinland, Géraldine Picaud, CEO of SGS, Stan Zurkiewicz, CEO of DEKRA, Chantal Guay, CEO of the Canadian Standards Council (SCC), Matt Gantley, CEO of the Royal Accreditation Service UKAS, Susan Taylor Martin, CEO of the British Standards Institution BSI, and Luc, Chief Scientist of Renault Group. Julia, as well as industry giants such as China Inspection and Quarantine Group and Huachuang Testing, jointly explored how to build a stable and trusted system against the backdrop of the rapid development of global AI.
TIC Summit
Summit Background:
Artificial intelligence is changing our lives. From optimizing decision-making processes to increasing productivity, the potential benefits are huge and undeniable.
To fully understand the impact of AI, we must consider its potential benefits, but also the ethical and social dilemmas it raises. Imagine a world where our decisions are influenced by AI and our freedoms are restricted by algorithms. If we fail to act responsibly and build appropriate safeguards, this dystopian scenario is not science fiction but a grim possibility. The future of AI lies not in what technology can do but in how we direct its potential.
Introduction to TIC International Inspection, Testing and Certification Council:
TIC International Inspection, Testing and Certification Council (TIC Council) is an international organization established in December 2018. It is formed by the merger of the former global inspection and certification industry organizations International Federation of Inspection Agencies (IFIA) and International Federation of Inspection and Certification Organizations (CEOC). The Council is headquartered in Brussels, Belgium, and has branches in the United States, India, and Shanghai, China. As the spokesperson for the global independent testing, inspection and certification industry, the Council advocates effective solutions to protect the public, support innovation and promote trade, continuously brings value to society, and ensures that consumers buy safe products, and makes best practices in safety, quality, health, ethics and sustainability. Its members come from more than 90 institutions and organizations in the inspection and certification industry around the world.
TIC International Inspection, Testing and Certification Council related information:
TIC International Inspection, Testing and Certification Council is a global trade association representing leading third-party testing, inspection and certification (TIC) companies, whose goal is to play a key role among standards bodies, IT manufacturers, national and international decision-making bodies and consumers to establish a common framework to design and manage trustworthy AI models. Artificial intelligence (AI) is reshaping the technology landscape.
Breakthroughs from self-driving cars to personalized healthcare systems are changing our lives and generating significant business value. However, AI systems are complex and dynamic, posing challenges to traditional conformity assessment. In addition, today we still lack standards and best practices for AI. Global collaboration is a must to achieve trust and mitigate risks.
● 1st from Left, Photo 1: SAIA Deputy Secretary General Xu Qi and TIC Council Global Secretary General Ms. Hannah Teddy
● Right Photo: SAIA Deputy Secretary General Xu Qi delivered a speech
● Right Photo From left to right in the picture are host Jennifer Baker, Alyssa Lefaivre Škopac of the Institute for Responsible Artificial Intelligence, and Xu Qi, Deputy Secretary General of SAIA
Xu Qi emphasized in the fireside dialogue that China has always paid great attention to and actively promoted AI ethics and governance and advocated international collaboration in the field of AI ethics and governance. He mentioned that China has also made significant progress in AI legislation. For example, in September 2022, Shanghai took the lead in issuing the "Shanghai Regulations on Promoting the Development of the Artificial Intelligence Industry" to promote the high-quality development of the artificial intelligence industry, strengthen the source function of new-generation artificial intelligence technology innovation, and promote the deep integration of artificial intelligence with the economy, life, urban governance and other fields. In July 2023, my country promulgated the "Interim Measures for the Management of Generative Artificial Intelligence Services", as the world's first regulatory legal document for generative artificial intelligence, which provides a strong guarantee for the healthy development of industry. On October 18 of the same year, China also issued the "Global Initiative on Artificial Intelligence Governance". China advocates that AI governance requires international cooperation, people-oriented, promotes AI for good, and advocates the establishment of a test and evaluation system for AI risk levels. We always adhere to the principles of fairness, avoiding bias, and avoiding differential treatment in processes such as data application, algorithm design, technology iteration, product development and application. Xu Qi then introduced the upcoming 2024 World Artificial Intelligence Conference (WAIC) to the guests at the TIC Summit and specifically mentioned that the 2024 WAIC will host a high-level meeting on global artificial intelligence governance, and sincerely invited colleagues around the world to pay close attention and participate. In addition, he also pointed out that the construction of the artificial intelligence governance system must be accelerated to keep up with the pace of AI innovation. This view resonated with the CEOs and executives of global testing, inspection, and certification agencies present, who expressed their common vision and determination to act quickly.
TIC Summit Golden quotes from the meeting
Dr.-Ing. Michael Fübi, Chairman of the TIC Board of Directors:
“AI as a technology is different in terms of conformity assessment, it’s different in terms of expectations, and we need to collaborate, not only within the quality ecosystem but also with society”
DEKRA CEO Stan Zurkiewicz :
“LLMs and ChatGPT are a revolution, and the speed at which they are upgraded breaks the boundaries. AI is certifiable and regulatable, and we should have the same responsibility as the ecosystem we have built in terms of quality management.”
UKAS CEO Matt Gantley:
“This is our golden opportunity to get ahead of the Gartner curve early in the process. We must collaborate on all aspects of quality infrastructure, develop skills and set standards.”
SGS CEO Géraldine Picaud:
“We need to upskill and train talent, but we also need to attract talent. We have to provide an ecosystem for these talents that is different from the traditional TIC, and we are doing that. ”
Alexander Hoefmans , Deputy Director, Digitalization, Privacy and Administrative Simplification, Cabinet Office of the Secretary of State, Belgium :
"I am quite familiar with the concept of trust. Trust, responsibility and transparency are the keys to digital transformation."
Frederico Oliveira da Silva, Senior Legal Officer, BEUC European Consumer Organization
“The responsible means respecting the rights of individuals and that AI must be placed on the market with defined safeguards. ”
London Mayor Michael Mainelli (Video Link):
“The AI genie has already been unleashed, and the assurance of recognition and certification is particularly important.”
Julia Luc, Chief Scientist, Renault Group:
"There is nothing in science that cannot be explained. AI is mathematics and can always be explained."
TIC International Inspection, Testing and Certification Council’s recommendations for AI governance:
1. Adopt a risk-based assessment framework to categorize systems according to their potential impact, risk, and use to facilitate their review.
2. Adopt robust security and security-by-design principles in AI system development.
3. Develop an AI testing framework or “AI certification framework” to allow for tailored assessments of AI systems based on their specific circumstances,
4. Determine the functional assurance and verification of the Al system to supplement the documentation requirements.
5. Establish a qualified third-party AI system, especially high-risk systems, and conduct independent qualification assessments.
6. Introduce lifecycle management of the Al system, including continuous monitoring of the reliability and fairness of the system, and re-evaluation during its lifecycle.
7. Respect the differences between horizontal and sectoral legislation and standardization between different industries, and transform the sectoral specificity of AI into the adaptability of standards.
Related initiatives issued by other countries and international organizations in the field of AI governance:
United Nations: In March 2024, the United Nations General Assembly adopted a non-binding resolution on artificial intelligence. The resolution aims to promote "safe, secure, and trustworthy artificial intelligence systems that will contribute to sustainable development for all."
USA: After passing the AI Bill of Rights in October 2022 and encouraging major AI companies to sign voluntary commitments, the United States published the Executive Order on Safe, Secure, and Trustworthy AI in October 2023. In February 2024, the White House established the American AI Safety Institute, whose mission is to "develop guidelines, evaluate models, and conduct foundational research to address the risks and opportunities of AI."
Canada: In 2017, Canada announced the Canadian Artificial Intelligence Strategy to strengthen its leadership and innovation in this field. The strategy focuses on three pillars, mainly commercialization, standards, talent and research. In addition, in September 2023, Canada launched a voluntary code of conduct on the development and management of advanced generative artificial intelligence systems, aimed at strengthening the accountability, security and transparency of artificial intelligence development. Currently, the Canadian House of Commons is considering the Artificial Intelligence and Data Act (AIDA) in committee. The bill was first introduced in June 2022 as part of the Digital Charter Implementation Act (DCIA). AIDA will introduce the concept of "high-impact systems" and require some artificial intelligence models to comply with strict security and transparency rules.
EU: The AI Bill was submitted in April 2021. Aiming to promote the development of trustworthy and human-centered AI, the AI Bill adopts a risk-based approach, prohibits a range of practices, and imposes strict obligations on high-risk practices. In December 2023, the co-legislators reached a provisional agreement, which was approved by the plenary session of the European Parliament in March 2024. The AI Bill was formally approved and is becoming a horizontal framework to address AI. It creates the Office of Artificial Intelligence to oversee general AI models and promote a consistent direction for Europe on the international stage. Clear obligations are also imposed on other high-risk AI systems. The European Parliament also proposed transparency requirements for general AI systems, including compliance with EU copyright law and publishing a detailed summary of the content used for training. Stronger GPAI general AI models that may pose systemic risks will face more requirements, including performing model evaluations, assessing and reducing systemic risks, and reporting incidents.
Organization for Economic Co-operation and Development (OECD): In May 2019, the OECD adopted its five AI principles, which aim to support a human-centered approach to trustworthy AI. The OECD’s principles formed the basis for the G20’s human-centered AI principles, which were adopted in Japan in June 2019.
G7:In October 2023, the G7 adopted guidelines and a voluntary code of conduct for organizations developing advanced AI systems, known as the Hiroshima Process. Based on a risk-based approach, it encourages regular reporting, information sharing, risk management, content certification mechanisms, and the quality of training data.
AI International Summit: In November 2023, the UK hosted the first AI Safety Summit. 28 participating countries signed the Bletchley Declaration, which emphasized the need for safe, human-centered, trustworthy and responsible design, development, deployment and use of artificial intelligence for the benefit of mankind.
