Geneva: The international community is turning its spotlight to the Swiss city of Geneva, which is hosting the Global Dialogue on AI Governance on July 6-7, 2026. The event convenes representatives from member states, governments, international organizations, the private sector, academic circles, and the technical community, reflecting growing international interest in establishing a more thoroughgoing and equitable global framework to regulate AI use and ensure that all nations benefit from the opportunities offered by this rapidly evolving technology.
According to Qatar News Agency, the dialogue is jointly organized by the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the International Telecommunication Union (ITU), in implementation of the mandate issued by the UN General Assembly. It is part of the commitments set out in the 'Global Digital Compact' adopted by the UN as a roadmap to strengthen international digital cooperation, including the development of global AI governance grounded in principles of fairness, transparency, inclusiveness, and respect for human rights.
The gathering is a consequential milestone in the path of global efforts to transition from theoretical debates on AI to the development of practical mechanisms for global cooperation, amid the unprecedented expansion of generative AI applications and tools. This is along with associated challenges related to cybersecurity, data protection and privacy, legal accountability, combating misinformation, and ensuring the responsible use of these technologies.
The UN affirms that AI holds significant potential to support economic and social development and improve services in sectors such as health, education, agriculture, energy, and transportation. However, these benefits won't be achieved equitably unless the global governance system is designed in a way that ensures all countries participate in crafting the policies and standards governing this technology and its applications.
According to the UN, participants are expected to explore mechanisms to strengthen international cooperation, exchange expertise, build capacities, and enable developing countries to benefit from AI applications, in addition to ways to reduce the widening digital divide and ensure that cutting-edge technologies reach different societies in a fair and inclusive manner. Among the key challenges facing the international community is how to strike a balance between encouraging innovation and accelerating the development of AI applications on one hand and establishing effective regulatory frameworks that limit potential risks on the other.
Participants are also expected to explore the development of shared principles for transparency and accountability, enhance trust in AI systems, and establish mechanisms for continuous risk assessment that keep up with the rapid evolution of this technology. In contrast, UN reports point to the need to address challenges related to energy consumption in data centers, reduce the environmental impact of the expansion of AI technologies, and ensure the sustainable use of digital resources.
According to UN sources, the dialogue marks increasing international and regional initiatives aimed at regulating AI use, as a wide swath of organizations and nations have launched legislative and ethical frameworks intended to promote responsible use of the technology. However, the multiplicity of initiatives prompts the need for greater international coordination to reduce discrepancies in standards and enhance global applicability.
Regarding the contribution of AI governance to setting international standards to reduce the risks of malicious cyber uses, Dr. Mohammed Saeed Al Saqatri, a technology and cybersecurity expert, told Qatar News Agency (QNA) that AI governance plays a significant role in establishing an international rules-based order that ensures the responsible use of this technology. This, he noted, comes through shared standards that promote transparency, accountability, and risk management, emphasizing that digital space knows no borders, and therefore the efforts of any single country won't be sufficient without genuine international cooperation.
Hence, initiatives led by various entities such as the UN, UNESCO, and ITU matter most in establishing rules that limit the misuse of AI in cyberattacks or in generating fake content and misleading people, Al Saqatri underlined. He stressed that to pull off this matter, there must be clear and practically enforceable international standards that define who is entitled to use AI models and with what level of authority, depending on the nature and risk level of each use case.
Al Saqatri added that these standards should be scrutinized and updated at least annually to keep up with the rapid advancement of AI capabilities, while ensuring that they aren't exploited unlawfully. He also emphasized the importance of establishing clear regulations governing how these models are trained and used, so that developers and operators are held accountable for their outputs.
In addition, Al Saqatri further pointed out that governance alone doesn't eliminate threats, but it is a necessary foundation for unifying policies, strengthening international cooperation, and enhancing confidence in these kinds of technologies. Regarding strengthening protective capabilities amid the increasing complexity of cyber threats, Dr. Al Saqatri said the cybersecurity landscape is changing rapidly due to AI, which now serves both defenders and attackers at the same time.
On one hand, it helps security teams detect threats more quickly, analyze large datasets, and respond to incidents more accurately. While on the other hand, he noted, attackers use it to develop smarter attacks, such as highly convincing phishing messages, malware that adapts to security systems, large-scale automated attacks, as well as deepfakes used in fraud and deception.
He continued that a clear example of this risk is what Google announced in May, when it discovered for the first time that a group of attackers had leveraged AI to identify a previously unknown zero-day vulnerability and write Python code to exploit it. The vulnerability was in the two-factor authentication (2FA) login system of a widely used website management tool, Al Saqatri explained, stating that Google detected the issue and patched the vulnerability before the attackers were able to exploit it at scale.
What revealed that the code had been written by AI was the presence of indicators within the code itself, such as instructional-style comments and illogical CVSS security rating- features commonly seen in outputs generated by large language models, Al Saqatri shared. Al Saqatri clarified that AI is also being used effectively on the defensive side, noting that a tool called AISLE achieved success in discovering 15 security vulnerabilities in a critical software library known as OpenSSL within a short period, including all twelve vulnerabilities disclosed by the company in a single security release.
He added that similar tools developed by Google have helped identify serious vulnerabilities in core systems that underpin the internet. Consequently, success in cybersecurity today is increasingly tied to those who use AI more effectively and more responsibly. Al Saqatri pointed to key security challenges facing governments and institutions amid the expanding use of AI applications, particularly in protecting the data used to train AI models.
This challenge, he suggested, became especially clear in 2025-2026 through real-world cases, including malicious code repositories known as "Basilisk Venom" that were embedded in training datasets, as well as another case involving the Grok 4 model, where attackers planted a hidden secret phrase across social media platforms that later entered the model's training data without anyone's knowledge.
He further noted the precipitous increase in the gravity of cybersecurity threats, often referred to as model theft, where attackers have been able to replicate near-complete versions of commercial AI models by sending millions of queries and analyzing their responses. This was achieved at a relatively low cost of only a few thousand dollars, compared to the millions spent by firms developing the original models, Al Saqatri noted, indicating that these risks are growing further with firms' increasing reliance on open-source software and libraries.
To address these challenges, he emphasized the importance of a thoroughgoing approach that includes updating laws, building security into systems from the earliest design stages, conducting periodic risk assessments, sharing security intelligence between entities, investing in training national talent, and leveraging AI-based solutions themselves in a carefully controlled defensive manner.
Al Saqatri stressed the need to strike a balance between leveraging AI's capabilities to strengthen cybersecurity and protecting privacy and personal data. This requires adopting the concept of "responsible AI", he underlined, where systems are designed and operated according to principles of privacy, security, and transparency from the outset, rather than as afterthoughts.
It also requires compliance with data protection regulations, minimizing data collection to what is strictly necessary, applying anonymization and encryption techniques, and subjecting high-risk systems to independent review and auditing, Al Saqatri suggested. Al Saqatri argued that clear governance frameworks and accountability mechanisms, along with cooperation between governments, the private sector, academic institutions, and international organizations, are essential to ensuring that AI's capabilities are used to strengthen cybersecurity while safeguarding individual rights, protecting privacy, and strengthening trust in these technologies.
On his part, Dr. Fath Al Aleem Ali Hija, a Professor of Cyber Defense Studies at the Joaan Bin Jassim Academy for Defense Studies in Qatar, said cybersecurity today is undergoing a tectonic shift that goes beyond traditional technological evolution due to the rapid advancement of AI technologies. Artificial intelligence, Ali Hija underscored, is no longer merely a tool for improving system efficiency; it has become a factor reshaping the cyber threat landscape per se and changing the relationship between attacker and defender, as well as the speed of attack development and response mechanisms.
From a defensive perspective, Ali Hija said, AI has enabled the development of advanced capabilities in big data analysis, anomaly detection, prediction of malicious behavior, and automation of monitoring and response operations. This has enhanced the efficiency of security operations centers, improved situational awareness, and reduced the time required to detect and respond to incidents, he suggested.
Ali Hija further stated that AI has also enabled attackers to develop a new generation of cyberattacks characterized by learning and adaptability, automation of attack stages, analysis of target environments, and the production of highly convincing fraudulent content, as well as continuous refinement of attacks based on defensive system responses.
He warned that the magnitude of AI governance stems from the fact that this technology is global in nature, while most legal and regulatory frameworks remain national. Therefore, the challenge is no longer limited to regulating technology within individual states but extends to building an international system