The year(s) of AI at MWC continued today, with focus shifting to geopolitical tensions and how to respond to the dominance of the US and China in key markets
Enabling Choice: Openness and Interoperability in Digital Ecosystems
Day 2 kicked off with a relatively quiet panel on hardware interoperability in digital ecosystems hosted by the Coalition for Open Digital Ecosystems (CODE). While a slightly smaller crowd could suggest that CODE’s message is both increasingly familiar and apparently successful in the context of recent regulatory interventions on digital competition, panellists nonetheless made a renewed case for the urgent importance of expanded and deepened interoperability in a range of devices. Coco Carmona (Executive Director, CODE) set the tone for a consumer-focused conversation, sharing that 82% of EU consumers (based on a survey of 5,000) say that interoperability of connected devices is important to them. Nicholas Porter (VP of MX Product Strategy and Commercial Europe, Samsung) echoed this message in describing how the consumer experience of seamlessly connected devices drove Samsung’s efforts to integrate standards-based interoperable functionality into their earliest stages of its development work. In addition to a range of well-cited consumer benefits of more open ecosystems – including more innovative and affordable devices and services – Erik Kay (VP of Engineering, Google) argued that open ecosystems also support more secure systems, denying the commonly argued trade-off between openness and security in a thinly veiled reference to Apple’s longtime position on the closed nature of their offerings. Through a more open environment, “more eyes” are able to find and correct issues that could pose privacy and security threats to consumers. Instead, Kay argued that the only trade-off is related to the speed of development for devices and services and suggested that consumers lose value in a number of ways when firms opt for closed systems that are launched more quickly. However, he did concede that openness is difficult in the EU when Member States often act in their own interests, although the value still outweighs the added effort. Porter offered a vision for the future of digital ecosystems that is responsive to consumer needs, including in relation to the expectation that consumers will keep devices for longer while still looking to capitalise on new and innovative services – a trend that will require interoperability built into systems today.
Keynote 8: Global Shifts
As President Trump’s administration prepares to implement tariffs to boost US-based AI chipmakers and reduce any reliance on foreign (particularly Chinese) developers, somewhat tense discussions at MWC25 delved into the current geopolitical landscape’s impact on the semiconductor market. These tensions surfaced between Keyu Jin (Global Economist, Harvard University) and Gregory C. Allen (Director, Wadhwani AI Center, Center for Strategic & International Studies), with the two disagreeing on the priorities for Chinese AI policy. Allen argued that the Chinese Government’s aim is to economically coerce the rest of the world into relying on its technology exports, particularly semiconductors. Jin played down these ideas and made the point that US actors are far more concerned about China than China is about the US – explaining that current Chinese policy efforts are focused much more domestically due to internal economic issues. More specifically, Jin said that there is a “dangerous” obsession with China in Washington D.C.. Allen persisted that this was the other way around, even claiming that Chinese military officials are the ones “obsessed” with the US, relaying his own concerns about China’s ability to “weaponise” the semiconductor supply chain.
The panel, which also featured Jerry Sheehan (Director, Directorate for Science, Technology and Innovation, OECD), did agree that the supply and value chains involved in the development of semiconductors needed to be far more diversified, avoiding concentrations at certain stages and from certain places, such as Taiwan. Jin questioned whether economic sanctions, including tariffs and other export controls, would be effective here, arguing that these have only contributed to more global fragmentation in technology development. Sheehan also raised the need to improve global coordination between OECD countries with those in the Global South to integrate AI effectively, helping less developed countries to climb the economic ladder. Sheehan was optimistic about how this would develop the Global South in the image of current OECD states – following their ideological and economic cultures.
Almost inevitably, the debate later returned to the future of AI development in China and the US. Allen was assured that the US’ tariffs and export controls were going to propel it ahead of China in the years to come. While praising DeepSeek’s achievements with its R1 model, developed at a fraction of the cost of most US models, Allen pointed out that larger US firms with far more capital and chips than DeepSeek could now use the Chinese firm’s methods to scale up their own operations, meaning that “Chinese AI is in trouble”. Jin shifted the discussion’s focus away from the US-China competition and instead highlighted how the two countries’ AI capabilities could actually complement one another, benefitting the greater good due to the differences in their focuses. Sheehan rounded out the panel by emphasising the significant economic potential of AI technology for the world, while warning that economic and national security are becoming inextricably linked, and will become more so due to AI.
AI under a Trump Government
Midday on Day 2 saw a well-attended session on the outlook for AI under the newly re-elected President Trump, with speakers acknowledging that the US Government’s approach, whether through public investment or protectionist policies, will shape the technology’s development globally and impact the AI ecosystem at large. Mallory Knodel (Executive Director, Social Web Foundation) described recent behaviour by the Trump Administration in this space as “careless”, including the roll back of President Biden’s AI Executive Order on system safety, which has left AI agencies in the country underfunded and under-resourced. Knodel added that there is a lack of political will to ensure international harmonisation of AI through standards, as reflected by certain countries – e.g. the UK and US – not signing the Paris AI Action Summit declaration. Mark Somol (CEO, Hyacinth AI) stated that there is no question that the US Government is squarely focused on “money and power”, which stands to benefit large domestic companies in the short term. He warned that regulation is something that could kill innovation, but that safety is a crucial nuance in the context of AI. Somol considered that safety is central to securing trust and accelerating the wide-scale adoption of AI, which therefore requires oversight and some guardrails to be in place. On the topic of regulation, Jose Ignacio Torreblanca (Senior Policy Fellow, European Council on Foreign Relations) was concerned by the US’ step away from efforts to establish common governance frameworks for AI, particularly following recent years of general consensus between the US and EU via the Trade and Technology Council. He argued that Europe should not be coerced into the geopolitical conflict between China and the US. Unfortunately, Somol foresaw heightened tensions between the two countries under the Trump Administration, as well as greater protectionism. He said that this would close off the value of the technology to citizens and stifle innovation, meaning that the US would only have set a trap for itself by doing so. Somol argued that resource constraints imposed on China led to the emergence of DeepSeek, which hit share prices of American firms and undermined recently announced US plans for massive private investment into AI infrastructure. Torreblanca felt DeepSeek represented an “embarrassment for Europe” where policymakers and industry had considered that such innovation could only be achieved by those with the deepest pockets.
AI Act and the GDPR: What it means for international cooperation and enforcement
In a very crowded conversation tucked in the very back corner of the venue, the EC presented its case to counter the emerging narrative that the AI Act and the GDPR are conflicting and unworkable. Meritxell Borràs Solé (Director, Catalan Data Protection Authority) acknowledged the complexity of the EU’s regulatory framework but cited the diversity of actors in the AI value chain as well as the cross-economy implications of the technology as the core causes of that complexity. Lieven Brouwers (Head of Data Flows Policy, DG Justice and Consumers) was direct and uncompromising in the idea that the AI Act is built upon the GDPR and that the two frameworks are fundamentally compatible in their shared assertion that consumers should have control over their personal data. Brouwers noted that the EU wasn’t unique in its efforts to grapple with the privacy implications of AI through regulatory intervention and mentioned ongoing investigations in Brazil, Japan and South Korea, as well complementary work underway within the G7 and G20. He did acknowledge that the differing remits of regulators across the bloc, particularly as Member States have designated different competent authorities to enforce the GDPR and AI Act, has complicated compliance for firms. However, he emphasised the EC’s plans to continue to try to support firms by offering guidance on best practices for AI adopters and to work with countries that have GDPR adequacy agreements with the EU (including Japan, South Korea and the UK) to maximise the benefits of these agreements and the value of international data flows as key assets.
Looking Skyward: NTNs and the Next Era of Global Connectivity
During a panel session on non-terrestrial networks (NTNs), representatives from the satellite industry were keen to outline their role in closing the coverage gap as an integral member of the evolving mobile ecosystem. Despite being relatively late on the day’s agenda, the session drew a strong crowd, reflecting rising interest in the sector as well as its growing presence at MWC. Isabelle Mauro (Director General, GSOA) highlighted the transformation of the industry over recent years, which has been underpinned by significant investment in multi-orbit launches and the provision of more affordable and reliable services for consumers. However, Parth Trivedi (CEO, Skylo) identified spectrum constraints as a key challenge to satellite providers’ continued progress, drawing parallels with the all too familiar concerns voiced by mobile operators. Carmel Ortiz (Senior Vice President, Technology & Innovation, Intelsat) echoed these thoughts, stating that regulators need to manage and balance competing demands for spectrum, which is a finite resource after all. She also called for “enlightened regulation” that does not over-reach or stymie technological and use case innovation. Reflecting on satellite players as global companies whose services do not respect borders, Eva Berneke (CEO, Eutelsat Group) discussed the difficulty of dealing with so many different regulatory frameworks and urged greater international harmonisation but noted that the necessary discussions to make that happen are yet to begin. Trivedi described direct-to-device (D2D) as “a new paradigm for regulators”, claiming to have responded to more consultations in the past five months than in the previous five years. He nevertheless saw this as encouraging as regulators are clearly looking to get to grips with the subject, following the lead of countries such as the US. David Goldman (Vice President, Satellite Policy, SpaceX) agreed that the “1.0” framework established in the US was a good starting point that has scope to be iterated and stated that it should serve as a useful reference point for regulators in other countries.
Elsewhere around the conference on Day 2
In a morning interview, Arthur Mensch (Co-Founder and CEO, Mistral AI) discussed the impact of AI on the future of the telecoms sector. When asked about his position on the regulatory burden in the EU, Mensch took issue with the emerging narrative among tech firms about the AI Act stating that while existing regulation is a barrier to investment and AI development, it is not the primary problem. Mensch instead pointed to the persistent shortcomings of the single market for the telecoms and AI sectors and a reluctance of EU firms to invest in new infrastructure as the key issues. When pressed on whether the EU could compete with the US and China in AI, Mensch promoted the use of open-source data across Member States and encouraged companies to integrate AI across their operations. Similar discussions about AI regulation and the use of open-source data went on throughout the day in Barcelona, with Apple co-founder, Steve Wozniak echoing Mensch’s comments in his speech at the Talent Arena event.
