Navigated

Europe must be ready when the AI bubble bursts

Marietje Schaake The writer is a fellow at Stanford University’s Institute for Human-Centered Artificial Intelligence. She is the author of ‘The Tech Coup’

Settings
Print
Share
Listen

One take away from the new US national security strategy is the extent to which Washington fears a strong EU, as a single market, a democratic bloc and, crucially, as a tech regulator. Meanwhile, Europeans often hear from Americans that they are losing the artificial intelligence race. The hyperscale model is presented as the sacred path, and Europe simply lacks the resources to compete.

But the resource-intensive AI platform bubble in which the US dominates cannot last. A market correction will shift attention to alternative models. This will in turn create new opportunities for Europe, which has strengths in applied AI and the chance to build the most-trusted AI stack in the world.

A German car manufacturer does not require a chatbot trained on the entire internet. It benefits from AI systems trained on high-quality engineering data to optimise manufacturing processes, predict maintenance needs or streamline safety reporting. A Dutch hospital needs diagnostic tools that meet medical standards, not general-purpose systems that may come up with medical disinformation. And a French bank needs AI that offers efficiency gains while adhering to strict financial services regulation.

As companies increasingly rely on AI-generated code, software vulnerabilities proliferate if security standards are missing. Rob Joyce, former cyber security director at the National Security Agency, has warned of “big wildfires of cyber burndown”. Jen Easterly, former director of the US Cybersecurity and Infrastructure Security Agency, imagines a world in which “secure by design” principles and zero-trust engineering will render the cyber security industry obsolete. When deployed well, AI can boost security and resilience. Europe should focus on building safety into systems from the ground up rather than patching vulnerabilities after deployment.

Rather than 27 member states pursuing their own goals, the EU needs specialised clusters and strategic ecosystem development that links universities, start-ups and investors. Estonia’s digital government infrastructure reminds us how small countries can carve out a position of leadership. Germany’s excellence in robotics and industrial automation is worldclass. French institutions produce frontier AI research. Europe should deepen such hubs of expertise and understand the leverage its chokepoints offer.

A coalition of the willing should demonstrate what European AI leadership looks like in practice. Beyond enforcing its laws, the EU should prioritise the fast-tracking of capital markets union, an increase in research funding and streamlined visa processes for tech talent. In the face of US hostility, the EU must be willing to play power politics.

AI factories represent one approach to democratising access by unlocking shared data and computing infrastructure for research and development. Yet more sovereign compute capacity and AI infrastructure is needed. Pooled resources can support data commons for research, AI systems for education and tools for democratic participation that serve the public interest. Having its own capacity should make the EU less vulnerable to geopolitical turmoil too.

Tech must be integral to plans for decoupling Europe from over-dependence on America in trade, finance and defence. While the EU debates what to do about AI platforms, those same platforms are embedded in critical infrastructure, government services and defence. That means a further transfer of knowledge and resources, and a loss of sovereignty.

But when the AI bubble bursts, valuations will reset. Talent will become available. Customers will question whether they need the most expensive, risky and least transparent systems. The NSS has unwittingly revealed that the Trump administration sees the EU’s strengths more clearly than Europeans do.

The US hyperscale model is not destiny. It emerged from a particular corporate culture with a high tolerance for risk, hands-off regulation, disregard for environmental harms, and a privileging of growth over other values. The EU should be confident about making different choices, in favour of trust, security, sector-specific excellence and democratic accountability. It must double down on developing an alternative before the next layer of dependencies becomes entrenched.

The question is not whether the AI bubble will burst, but if Europe will seize the moment when it does.

Europe must be ready when the AI bubble bursts

Comments

Post a Comment

Popular posts from this blog