International dialogue has taken center stage in recent weeks, with a series of important gatherings taking place in quick succession. The Shanghai Cooperation Organization (SCO) meeting, the BRICS Summit, and the UN General Assembly all attracted significant attention this year, coming as they did amid a series of cascading crises that have continued to intensify.

What was particularly notable about these events was the growing confidence of voices and coalitions from the Global South, which increasingly appear to be among the strongest advocates of multilateralism in the present moment. This was evident in the unexpectedly strong support for Palestine and Iran expressed in the Bishkek and New Delhi Declarations. It was also apparent, however, across a much broader range of issues.

In the world of technology, for instance, both the SCO and BRICS summits featured significant calls for the development of global frameworks for AI, alongside announcements of South-South collaborations to advance this agenda. At the UN General Assembly, meanwhile, the global majority called for a greater role in shaping the rules of the digital economy. The issues raised spanned the implementation of the Global Digital Compact, access to digital infrastructure, and the regulation of critical mining industries that provide the material foundations of frontier technologies. This increasingly assertive vision is a promising development at a time when the US appears intent on undermining international cooperation in pursuit of its own interests.

The latter’s strategy, however, appears to be going astray on the question of AI, with Trump’s statements at the General Assembly running counter to the direction in which Silicon Valley’s leading companies have been moving. Indeed, the past month has seen a heated debate following the resignation of an Anthropic employee who became a whistleblower, raising concerns about an existential threat to humanity stemming from the current trajectory of AI development. His warnings quickly went viral, drawing support from several others who continue to work within the industry.

Soon after, Anthropic CEO Dario Amodei released a statement calling for a collective slowdown in the development of frontier models. Surprisingly, the statement was subsequently endorsed by Elon Musk, Sam Altman, and Demis Hassabis, producing a rare consensus among the most significant figures in the American AI industry.

While it is certainly true that the current state of the technology has raised alarming concerns — from cybersecurity risks and potential abuse by malicious actors to mass economic disruption and an exponential proliferation of misinformation — many have nevertheless viewed this series of events with incredulity. Questions have been raised about the vagueness and staged nature of the initial whistleblower episode, the sudden convergence of opinion among bitter rivals, and the suggestion that these tech billionaires are motivated by genuine concern for the public interest.

Moreover, given the growing threat posed by open-source alternatives, the difficult path to IPOs for many of these companies, and increasing scrutiny of a potential financial bubble in the industry, some have speculated that the supposed “slowdown” may have been orchestrated to serve other objectives. Indeed, it is not even clear whether any substantive pause has actually been implemented in the development of new models.

All that said, the stock markets and Trump reacted negatively to the news, with the latter immediately declaring the strategy a non-starter and reiterating that “winning on AI” was a geopolitical priority for his administration. Given this apparent misalignment between the state and its national champions, it is possible that neither will entirely get what it wants, while both continue to aggravate their mutual frustrations.

Meanwhile, data center projects are being derailed across the globe amid an increasingly volatile economic climate. This month, Oracle announced that it was invoking a force majeure clause due to serious delays to one of its major construction projects. At the same time, a number of data centers in West Asia have been caught in the crossfire of war and remain vulnerable to deliberate destruction.

Beyond this, mass protests against data center infrastructure are becoming increasingly pronounced, with hundreds of projects stalled in the US and new movements emerging across the Global South. These developments point to growing public disaffection and a demand for much greater oversight. It is crucial that the international community seize this momentum rather than allow Silicon Valley to shape the regulatory frameworks that the industry increasingly requires.

In other news, at the end of August, Meta moved to settle a landmark lawsuit over social media addiction, agreeing to pay $18 billion in damages and implement a series of reforms to its platforms. The settlement comes on the heels of multiple cases the company has lost this year, all of which alleged that the design of its algorithms had caused significant harm to users. While critics have argued that the settlement allowed the company to avoid more radical changes, such legal action is likely to continue pushing for further changes. It is an opportune time for policymakers to think about what kind of restrictions on algorithms may have the most far-reaching societal impact.

This development also coincides with a growing policy consensus around restricting social media use by minors, with dozens of countries across Europe and Asia introducing legislation to this effect. While this demonstrates a welcome appetite for legislative intervention, such bans do not address the root of the problem. As surveys of young people’s opinions also testify, the issue lies in the design of these platforms themselves and, ultimately, in the underlying political economy that shapes their incentives and decision-making. Only by making these issues central to our regulatory agenda will we be able to forge a path towards renewing our public spheres.

The Sins & Synergies Lounge

Don’t miss this timely word of caution from Brian Chen on the future of the fight against data centers and the emergence of a new structure of accommodation that is taking shape.

Also read this moving account of what it means to lose Autistici/Inventati, the independent Italian tech collective whose services have supported activists and alternative media for 25 years.

Put the Basel Action Network’s report on the AI e-waste crisis on your radar, which looks beyond the energy and water costs of AI to examine its rapidly growing material footprint.

For a broader look at the material infrastructures behind AI, explore this edited collection bringing together perspectives on the environmental, social, and political dimensions of the systems that make AI possible.

Tune into this critical conversation at the launch of the Digital Hate Review, a new journal that pays attention to how online hate has evolved over time and how the issue touches public policy, content moderation and platform policy, AI governance, and a variety of other topics at the intersection of technology and public discourse.

Finally, read Geert Lovink’s farewell lecture, which offers a sharp reflection on what alternative strategies can still do in an age of platform domination if we start using the media we’ve created as tools, not as environments that cannot be turned off.