Artificial intelligence (AI) permeates every aspect of our lives, from everyday communications to global security systems. The interest of leading countries in AI issues in this era of digital transformation is no coincidence but reflects the growing awareness that AI is not only a technological breakthrough, but also a catalyst for change in many areas of social and economic life. This is evidenced by recent events such as the AI Safety Summit or the resolutions on advanced AI systems adopted by the G7 in Hiroshima. This trend is also confirmed by intensive regulatory work, including the AI ACT, which has been in the works for some time.
World leaders speak with one voice on the risks of artificial intelligence
The G7 recently published two documents aimed at harmonising the approach to advanced AI systems: the ‘Guiding Principles for Organisations Developing Advanced AI Systems’ and the ‘Code of Conduct for Organisations Developing Advanced AI Systems’.
The position developed by the Group is based on eleven fundamental principles, which emphasise:
- The need to identify and mitigate risks associated with AI
- Transparency of AI vulnerabilities
- Incident reporting
- Investment in cybersecurity; and
- Protection of personal data and intellectual property
At the beginning of November, the AI Safety Summit 2023 took place in the UK, bringing together politicians, representatives of leading AI companies, social organisations and researchers from nearly 30 countries in Europe and around the world.
Experts discussed the risks of artificial intelligence and how to effectively mitigate their impact through coordinated international action.
Summit participants also adopted the so-called Bletchley Declaration, which underlines the urgent need to identify and address AI threats. The document notes the importance of security, transparency, accountability and privacy in relation to AI. The declaration also highlights the need for international cooperation to minimise risks and harness the potential of AI for the public good, in line with the UN Sustainable Development Goals.
The organisers, who are particularly keen to develop global dialogue and concrete work on AI security, are announcing further meetings and events on the topic.
European Union announces regulation of artificial intelligence
Meanwhile, work on the European Union’s AI regulation, known as the AI Act, is approaching a critical point.
Following recent negotiations, which focused on the issue of classifying high-risk AI systems, a further meeting is planned for December to finalise the shape of the regulation.
Discussions are expected to include the use of AI by law enforcement, prohibited practices in the use of AI, and potential impacts on fundamental human rights and the environment.
The issue of remote biometric identification is controversial and the positions of the various negotiating parties continue to diverge.
What aspects of AI need special attention
Looking at the work on AI ACT to date, several key areas emerge, such as:
- Accountability of decisions made by AI systems
- Transparency and explainability of algorithms
- Data protection and privacy
More recently, however, the issue of security has come to the fore, as reflected in an executive order issued by US President Joe Biden.
The order includes a call for the creation of new security standards, requiring companies developing AI systems to share the results of security testing with the US government.
The decisions that will be made in the coming weeks will have a significant impact on the global approach to the technology and the future direction of AI.
The Bletchley Declaration, the findings of the G7 and the US government’s order show that security is an integral part of the innovation process, and the most important aspect to emerge from the work and discussions on AI is the need for a balanced approach to new technologies. One that promotes the development and implementation of novelty on the one hand, and addresses security and its ethical, social and legal aspects on the other.
Questions? Contact us