The forum began with a speech highlighting a global push towards regulatory simplification, including efforts in the US, to remove outdated rules and accelerate approvals. Spectrum policy, AI governance and supply-chain resilience were also central themes, each carrying geopolitical and economic consequences. The speaker noted that US leadership depends on partnership, technical rigour and the willingness to modernise, echoing the view, articulated at the International Telecommunication Union, that regulators are builders of the global digital ecosystem.
Spectrum policy
This session began by linking the issues of global coordination, connectivity gaps, WRC-27 priorities and the dual role of AI as a driver of demand and an efficiency tool. A panellist argued that the United States is at a pivotal moment: Congress has restored the FCC’s auction authority and set an ambitious target to free up 800 MHz of spectrum. The challenge, he said, now shifts from theory to delivery: sequencing bands, aligning stakeholders and meeting legally mandated timelines, including a major C-band auction.
There was a need for both licensed and unlicensed spectrum, but the core test is ‘highest-value use’ and consumer outcomes. Fixed wireless access growth is evidence that mid-band auctions have enabled real competition in broadband, including against cable incumbents.
Another contributor argued that the future is coexistence: satellite for remote coverage, mobile for mobility, and fixed wireless and unlicensed/Wi-Fi as the dominant indoor access layer. He warned that AI will shift traffic patterns towards more upload and symmetric demand (e.g., video inferencing), increasing the need for wider channels across both licensed and unlicensed spectrum. A panellist from a network operator echoed the importance of global harmonisation and economies of scale, urging sustained momentum on mid-band pipelines so operators can invest on predictable timelines.
Transatlantic update
This session explored the shifting transatlantic technology agenda in light of the EU’s new ‘digital omnibus’ simplification package and its implications for AI, data and investment. The current EU-US baseline agreement rules out ‘fair share’ network usage fees, includes a major EU commitment to purchase AI chips (with export security assurances under discussion), advances mutual recognition linked to the EU Cyber Resilience Act and prioritises standards work, especially 6G, while avoiding EU digital regulation such as the Digital Markets Act and Digital Services Act.
The digital omnibus was presented as an early step in a wider burden-reduction drive: consolidating data legislation around the Data Act, clarifying GDPR concepts relevant to AI training (including pseudonymisation), extending timelines for high-risk AI obligations to allow standards to mature and streamlining breach reporting into a one-stop process. On convergence with the US, the speaker anticipated more practical alignment in industry practice than in law, while reaffirming support for transatlantic data flows.
Cybersecurity and critical infrastructure protection
The session began with the moderator stressing that cyber risk is now inseparable from the rapid expansion of digital connectivity. As networks extend to drones, IoT, vehicles and critical infrastructure, ‘one vulnerability is a vulnerability for all’.
A panellist described a ‘hyper-connected’ world with a larger attack surface and heightened geopolitical incentives for adversaries, but also argued that the same technological shift, particularly AI, creates new opportunities for detection, automation and ‘security by design’, including embedding stronger security as 6G standards take shape. He emphasised that trusted infrastructure must be paired with stronger cyber practices, information sharing and capacity building with partners.
It was noted that trade agreements increasingly reference ‘trusted networks’ but it wasn’t clear if this was according to technical security or also geopolitical risk. A panellist argued that it should include security-by-design and zero-trust architectures, plus robust information-sharing frameworks that don’t over penalise operators. Panellists warned of national-security exposure, economic competitiveness and a threat to democratic values if authoritarian models of control and censorship are exported through technology stacks.
Deep dive: Latin America
This session explored how regulators, policymakers and development institutions in Latin America and the Caribbean are adapting to rapid technological change across connectivity, media, data and AI.
Satellite
The growth of low-earth-orbit constellations highlights concerns around orbital congestion, spectrum interference and long-term sustainability. For large and geographically diverse countries, satellites were positioned not as a single solution but as part of a broader technology mix alongside fibre and mobile networks. Concrete examples illustrated how remote communities, particularly in riverine and rural regions, depend on satellite and hybrid solutions to access healthcare, education and public services, reinforcing the idea of ‘meaningful connectivity’ rather than coverage alone.
Audiovisual
Efforts to modernise decades of television rules into a coherent framework aligned with a converged, platform-driven environment are continuing. Central themes included pluralism, audience participation and safeguards for children and vulnerable users. A newly adopted child online safety law was presented as a move from fragmented interventions to a whole-of-state approach, emphasising prevention, education and safety by design.
Investment
Traditional last-mile connectivity projects remain important, but demand is growing for more flexible tools including reforming universal service funds, channelling finance through public development banks to smaller local providers and using policy-based loans tied to regulatory reform. These approaches can translate technical advice into scalable programmes providing real benefits for end users.
Spectrum
On spectrum policy, the panel underlined the need to balance global alignment with national priorities. Upcoming low-band awards are essential for rural coverage, while policy shifts in mid-band spectrum reflect changing views on the balance between unlicensed use and mobile broadband needs.
Finally, the panel examined adoption gaps, resilience and future infrastructure, including affordability programmes that dramatically increased uptake once cost barriers were addressed. Data centres, subsea cables and disaster resilience featured prominently, particularly for smaller states.
Keynote: digital sovereignty
A senior executive at a global payments and technology company noted that digital sovereignty is now a permanent feature of technology policy. She suggested that the concept has moved from the margins into mainstream policymaking, particularly as the US and EU debate competition, data governance and resilience in the digital economy.
Digital sovereignty remains an ambiguous term but, she argued that this creates a window to shape what it means in practice. She defined sovereignty as control and went on to describe how governments seek control across different layers of the digital stack: data, infrastructure operations and, in its most extreme form, technology independence. Governments feel a sense of urgency given the concentration of compute (the processing power in data centres), data and models in a small number of countries.
Drawing on the payments sector – often treated as critical infrastructure – she described how states have increasingly pursued state-backed digital public infrastructure to reduce reliance on external providers. However, she warned against maximalist sovereignty approaches that drive fragmentation, reduce interoperability and weaken cybersecurity cooperation.
Internet governance
This panel discussed the WSIS+20 (World Summit on the Information Society) negotiations then unfolding in New York and asked: is the Internet Governance Forum (IGF) genuinely at risk, and what happens if governments fail to agree?
One speaker warned that the IGF has effectively been ‘on life support’ for years. Even if its mandate is renewed, funding and political commitment have steadily weakened and a vote-based outcome could trigger chaos, invite unrelated geopolitical issues and accelerate fragmentation. Others echoed that the greatest strategic risk is the erosion of the multi-stakeholder approach. If the process fails, sceptical governments could claim the model is unworkable and shift back to narrower, state-only forums.
Speakers argued that the WSIS remains the key UN framework that recognises distinct roles for different stakeholder groups, including the technical community. They highlighted the growing importance of national and regional IGFs.
Participants also underlined that the toughest debates are less about principles than financial mechanisms: not only how to fund the IGF, but how to resource the broader WSIS architecture without proliferating new processes.
Fireside: international engagement
The second day opened with a dialogue between a senior executive from a major US telecommunications operator and a former senior government official. The discussion explored why international engagement remains essential for a largely domestic carrier, emphasising that networks, standards, spectrum and trusted supply chains are inherently global and that security and resilience depend on alignment with international partners.
The speaker described how AI is already embedded across network management, cybersecurity and customer service, delivering efficiency and resilience. At the same time, she warned that fragmented regulation, particularly a growing patchwork of US state-level AI laws, risks undermining innovation, arguing strongly for a coherent federal framework and alignment with international approaches, especially in Europe.
The conversation highlighted AI’s capital intensity. While core AI infrastructure is likely to consolidate among a small number of players, innovation at the application layer is expected to remain dynamic and competitive.
While AI and data centres may experience cyclical corrections, AI can be seen as the catalyst that allows societies to fully convert digital information into sustained productivity and economic growth.
Subsea and fibre
A senior policy lead from a major global technology platform explained why network infrastructure is a critical, and often overlooked, pillar of AI. Alongside data and compute , AI depends on high-capacity connectivity, subsea and terrestrial, to move data between facilities and users. The speaker highlighted the scale of recent fibre and subsea investments, including large multi-partner builds and new projects designed for resilience and global reach.
It was argued that the biggest constraint is not engineering but policy. Fragmented permitting across jurisdictions, slow approvals, restrictive maritime and customs rules and weak regulatory predictability risk delaying projects and stranding investment. In the US, problems arise from opaque national security review processes and short-term permits that clash with decades-long asset lifecycles.
Infrastructure
This session drew on two recent US Department of Commerce consultations to explore what an effective AI infrastructure policy might look like in practice.
A senior official outlined a new initiative focused on ‘exporting the American AI stack’, emphasising that the programme is still being defined and depends on public input. Key questions include what counts as the AI ‘stack’, how consortia should be structured and which markets the US should target. A former national security adviser argued that rivals have succeeded with cheap, turnkey technology packages, so the US needs tailored offerings matched to each country’s needs and paired with credible financing to provide a viable alternative and strengthen both economic and national security.
A cloud and security executive pushed for cybersecurity to be foundational, not bolted on, and argued that the export concept must extend beyond training infrastructure to include inference and deployment at the edge. On ‘sovereign AI’, the panel favoured a definition rooted in freedom of choice – control of data and vendor flexibility – rather than rigid data localisation. ‘Sovereignty’ can become a non-tariff barrier that fragments interoperability across allied ecosystems.
Online safety
A guest speaker traced the field’s evolution from the mid-1990s ‘dawn of the web’, when the dominant concern was harmful content. In the mid-2000s and with Web 2.0, the risks shifted from static content to behaviours: social networking, cyberbullying, sexting, overuse and oversharing.
It was argued that online safety has now entered a third phase, bringing new opportunities but also multiple streams of risk. Three stood out, the speaker noted: children forming emotional attachments to chatbots that tend to mirror and reinforce user views; the erosion of trust as synthetic media and deepfakes make it harder to believe what is seen and heard; and the potential weakening of critical thinking as students rely on AI tools to complete schoolwork, undermining ‘learning how to learn’.
Both speakers considered a recent national approach overseas restricting young teens’ access to social media as highly controversial, with emerging unintended consequences including in child rights. They favoured more targeted restrictions, particularly limiting phone use in schools, and stressed that parental fears in emergencies can be a significant barrier to adoption. They also emphasised practical harm reduction: delaying smartphones for younger children, using simpler devices where needed and prioritising values and ‘right from wrong’ before giving a child full internet access.
From policy to performance at the Department of Defense
The final session of the forum focused on the challenge of turning policy intent into operational reality, drawing on lessons from managing a large, complex public-sector IT portfolio.
Using defence IT modernisation as an example, the speaker argued that many failures stem not from poor intent but from weak governance, fragmented authority and limited visibility over spend and performance. In large bureaucracies, responsibility for outcomes is often separated from control of budgets, making coalitions and cross-organisational partnerships essential to drive change. Progress depends less on issuing new standards and more on enforcing existing ones, backed by credible data.
Interoperability was highlighted as a persistent challenge, particularly in multinational and multi-service environments. Attempts to ‘bolt on’ solutions late in the process were contrasted with more successful approaches that brought users, engineers, security teams, partners and vendors together early to design shared architectures.
Policies succeed when leaders actively manage accountability, stakeholder alignment and feedback, recognising that change is constant and execution is as important as strategy.
This reported was drafted using ChatGPT.
and then