G7 online safety agreement puts children and safe AI growth at centre of digital policy

The Group of Seven (G7) Digital Ministers have agreed their first common approach to protecting children and young people from online harm, marking a new step in international digital safety coordination. The agreement, reached during G7 Digital Ministers’ talks in Paris on 29 May 2026, covers children’s online safety, age assurance, artificial intelligence chatbot risks, digital literacy, platform accountability and safer artificial intelligence growth. The United Kingdom and its G7 partners also agreed measures to support small and medium-sized enterprises in adopting artificial intelligence through a tool developed with the Organisation for Economic Co-operation and Development. The agreement matters because digital regulation is now moving beyond domestic lawmaking into shared standards across major advanced economies, where children’s safety, artificial intelligence trust and business adoption are becoming part of the same policy conversation.

The Department for Science, Innovation and Technology said the G7 agreement sets expectations that children’s safety should be built into digital services from the start rather than treated as an afterthought. The agreement also calls for effective age assurance and closer cooperation between digital service providers, children, parents and guardians. G7 countries also agreed that data sharing between online platforms, parents and researchers should improve so the impact of digital services on children’s wellbeing can be better understood. The commitments come as the United Kingdom prepares to respond to its consultation on protecting children from online harms, which included potential bans or curfews for under-16s, restrictions on harmful app features such as infinite scrolling, and stronger parental controls.

Why does the G7 online safety agreement matter for children and young people in digital spaces?

The G7 online safety agreement matters because it turns online child protection into a shared international policy priority rather than a set of disconnected national debates. Children and young people use platforms, games, messaging apps, artificial intelligence chatbots and content services that operate across borders. A harmful design feature or weak age assurance system in one market can affect users in many others. That is why a common G7 approach carries more weight than a single national statement.

The agreement focuses on risks that have become central to digital childhood: harmful content, exploitation, weak age checks, low digital literacy, opaque platform design and emerging risks from artificial intelligence chatbots. The inclusion of artificial intelligence chatbots is especially important because children are increasingly interacting with systems that can simulate conversation, respond emotionally and offer information that may not always be safe, accurate or age-appropriate. In regulatory terms, artificial intelligence is no longer just a productivity tool or industrial technology. It is also becoming part of the everyday digital environment in which children grow up.

For governments, the policy challenge is to protect children without cutting them off from legitimate learning, communication, creativity and social interaction. That balance is hard because online services change quickly, business models reward engagement, and young users are often early adopters of new tools. The G7 agreement does not solve those issues by itself, but it creates a shared baseline: digital services used by children should be designed with safety, age suitability and accountability in mind from the beginning.

How does the G7 agreement link age assurance, digital literacy and platform accountability?

The G7 agreement links age assurance, digital literacy and platform accountability because each element addresses a different part of the online safety problem. Age assurance aims to ensure that children do not access services or features that are inappropriate for their age. Digital literacy helps children, parents and guardians understand online risks, misleading content and safe behaviour. Platform accountability places responsibility on digital service providers to design and operate services in ways that reduce foreseeable harm.

The age assurance element is one of the most sensitive. Governments want stronger checks, but age assurance must also respect privacy, security and proportionality. Poorly designed systems can collect too much data or create exclusion risks. Weak systems, however, can leave children exposed to adult content, exploitative contact, addictive features or harmful algorithmic recommendations. The G7’s emphasis on effective age assurance suggests that major economies are converging around the view that voluntary, easily bypassed age gates are no longer enough.

Digital literacy is equally important because technical controls cannot manage every risk. Children, parents and guardians need better tools to recognise harmful content, deceptive material, online manipulation and artificial intelligence-generated media. At the same time, the burden cannot sit only with families. The agreement’s focus on providers shows that the G7 expects digital service companies to build safety into products rather than simply telling parents to “monitor usage,” which is the online safety equivalent of handing someone an umbrella during a hurricane and calling it weather policy.

Why are artificial intelligence chatbots now part of the child online safety debate?

Artificial intelligence chatbots are now part of the child online safety debate because conversational artificial intelligence can blur the line between tool, companion, tutor, search engine and entertainment product. Children may use artificial intelligence systems for homework, advice, creative play, emotional support or casual conversation. That makes safety risks more complex than conventional content moderation because the system is not only displaying content, it is generating responses in real time.

The G7 agreement identifies risks to children from artificial intelligence chatbots as part of its shared principles. That inclusion reflects concern that artificial intelligence systems can produce inappropriate content, reinforce harmful behaviour, provide inaccurate guidance or interact with vulnerable users in ways that are difficult to supervise. The challenge is not limited to one platform or one country. Artificial intelligence models can be deployed globally, integrated into many services and accessed through tools that children may use without fully understanding what they are interacting with.

For policymakers, the chatbot question also raises accountability issues. If a child receives harmful advice or unsafe content from an artificial intelligence chatbot, responsibility may be spread across the model developer, platform provider, app distributor, school environment, parent controls and regulatory framework. The G7 agreement does not settle that accountability chain, but it signals that the world’s major advanced economies now see children’s interaction with artificial intelligence as a core safety issue, not a future problem to be filed under “deal with later.”

How does the G7’s safe artificial intelligence agenda connect trust, security and economic growth?

The G7’s safe artificial intelligence agenda connects trust, security and economic growth because artificial intelligence adoption depends on public confidence as well as technical capability. G7 countries reaffirmed their commitment to ensuring that artificial intelligence is developed and used in ways that people can trust, while recognising the technology’s potential to support economic growth and improve everyday life. That dual framing matters because governments are trying to promote artificial intelligence adoption without ignoring risks around misuse, cyberattacks, harmful content, deception and vulnerabilities.

The Paris talks also highlighted threats such as cyberattacks and the development of chemical and biological capabilities. Those risks place artificial intelligence policy inside the national security agenda. Artificial intelligence can support innovation, scientific discovery and productivity, but it can also lower barriers for harmful actors if safeguards are weak. The G7 discussion under France’s Presidency included further work on a mutual understanding of artificial intelligence risk assessment frameworks, which points toward more structured cooperation on how governments and institutions evaluate artificial intelligence risk.

The economic growth component is visible in the commitment to support small and medium-sized enterprises. Large technology companies have the capital, staff and infrastructure to experiment with artificial intelligence quickly. Smaller businesses often lack those resources. By supporting small and medium-sized enterprises through an Organisation for Economic Co-operation and Development-linked tool, the G7 is trying to prevent artificial intelligence adoption from becoming a privilege reserved for deep-pocketed companies. That matters for productivity, jobs and competitiveness across advanced economies.

Why does the Organisation for Economic Co-operation and Development tool matter for small businesses?

The Organisation for Economic Co-operation and Development tool matters because artificial intelligence adoption is not simply about whether a business wants to use new technology. Small and medium-sized enterprises need to understand their readiness, skills gaps, data practices, workforce knowledge and operational use cases. A tool that helps small businesses assess artificial intelligence readiness can turn artificial intelligence from a vague strategic aspiration into a more practical business planning exercise.

The G7’s decision to connect artificial intelligence adoption with small business capability is important because small and medium-sized enterprises form a large part of employment and economic activity across G7 economies. If artificial intelligence benefits concentrate only among major technology firms or large corporations, the productivity gains could be uneven. Supporting smaller companies to adopt artificial intelligence responsibly helps broaden the economic impact of the technology.

There is also a safety dimension. Businesses that adopt artificial intelligence without understanding data protection, security, intellectual property and workforce implications may create avoidable risks. The G7 approach suggests that adoption and responsibility should move together. That is a healthier model than treating artificial intelligence as a magic spreadsheet fairy that arrives, fixes productivity and leaves no compliance paperwork behind. In reality, small businesses need practical guidance before artificial intelligence tools become embedded in operations.

How does the G7 approach address AI-generated content, data flows and digital infrastructure pressure?

The G7 agreement also addresses detection of artificial intelligence-generated content, trust in data and the resilience of the digital and artificial intelligence sector. Ministers highlighted the importance of improving detection of artificial intelligence-generated content so users, including children, can identify misleading or deceptive material online. That is directly relevant to misinformation, impersonation, fraud, deepfakes and online manipulation.

The agreement also reiterated the importance of trust in data as a foundation for innovation, with a commitment to enable cross-border data flows while maintaining strong protections for privacy, security and intellectual property. That balance is central to digital trade. Data must move across borders for cloud services, artificial intelligence development, research, business operations and scientific cooperation, but unmanaged data flows can increase risks around privacy, surveillance, misuse and intellectual property leakage.

Digital infrastructure pressure is another important part of the agreement. G7 ministers recognised growing pressure on energy and infrastructure as artificial intelligence adoption increases, while also noting that artificial intelligence and wider digital technologies can help improve efficiency and optimise energy systems. This is a practical issue because artificial intelligence growth requires data centres, chips, electricity, cooling, connectivity and resilient networks. Safe artificial intelligence policy therefore cannot stop at model rules. It must also confront the physical infrastructure beneath the digital economy.

What does the G7 agreement signal about the future of global digital regulation?

The G7 agreement signals that global digital regulation is moving toward shared principles even when countries keep different legal systems. The United Kingdom, Canada, France, Germany, Italy, Japan, the United States and the European Union do not regulate digital platforms in identical ways. However, the agreement shows that major advanced economies are trying to align around common expectations for children’s safety, artificial intelligence trust, data sharing and responsible innovation.

This matters because digital platforms and artificial intelligence developers operate internationally. If regulatory expectations diverge too sharply, companies face complexity, regulators face enforcement gaps and users face uneven protection. Shared principles help create a common floor, even if domestic enforcement continues through national laws such as the United Kingdom’s online safety regime and European Union digital regulation.

The agreement also suggests that digital policy is now a mainstream geopolitical issue. Child protection, artificial intelligence safety, data flows, digital infrastructure and small business productivity are no longer separate technical conversations. They sit inside economic security, democratic resilience and public trust. The G7 is not a world government, and nobody should pretend it can regulate the internet by communiqué. But when the world’s major advanced economies agree the same direction of travel, platform boards and artificial intelligence developers usually notice.

What are the key takeaways from the G7 online safety and safe AI growth agreement?

• The G7 Digital Ministers agreed their first common approach to protecting children and young people from online harm during talks in Paris on 29 May 2026. The agreement covers harmful content, exploitation, digital literacy, artificial intelligence chatbot risks, platform accountability and stronger safety expectations for digital services used by children.
• The agreement sets an expectation that children’s safety should be built into digital services from the start and supported by effective age assurance. The G7 approach also calls for closer cooperation between digital service providers, children, parents and guardians so online safety is not treated as a late-stage compliance add-on.
• G7 countries agreed that data sharing between online platforms, parents and researchers should improve to better understand how digital services affect children’s wellbeing. That commitment is important because regulators and families often lack clear, comparable evidence about the real-world impact of platform design on young users.
• The agreement follows the closing of the United Kingdom’s consultation on protecting children from online harms, which sought views on measures including possible under-16 bans or curfews, restrictions on harmful app features such as infinite scrolling and stronger parental controls. The United Kingdom Government plans to respond soon.
• G7 ministers reaffirmed the need for artificial intelligence to be developed and used in ways that people can trust, while recognising the technology’s potential to support growth and improve daily life. The talks also covered threats including cyberattacks and chemical and biological capability risks linked to artificial intelligence misuse.
• Small and medium-sized enterprises are expected to receive support for artificial intelligence adoption through a tool developed with the Organisation for Economic Co-operation and Development. The tool is intended to help smaller businesses assess artificial intelligence readiness, workforce knowledge and areas where adoption can be improved.
• G7 ministers agreed on a Vision on AI Openness and recognised that artificial intelligence models can support innovation, scientific discovery and economic growth. The agreement also emphasised the need to improve detection of artificial intelligence-generated content, particularly where misleading or deceptive material could affect users including children.
• The agreement also covered cross-border data flows, privacy, security, intellectual property and digital sector resilience. Ministers recognised growing pressure on energy and infrastructure from artificial intelligence adoption while noting that artificial intelligence and wider digital technologies can also help optimise energy systems and improve efficiency.