Whit Europe has been awa frae the grand AI safety debate

Europe’s dilemma ower AI wis set oot in stark terms this week. The heid o the continent’s central bank, Christine Lagarde, said Europeans hae twa options: shun the technology and tine the chance o growth; or embrace it and come tae depend on tools biggit in the US and China.

The great debate anent AI safety that has eruptit lately thraitens tae mak the choice moot. Gin the warst scenarios come tae pass – and experts hae varyin views on this – then the technology will impact the continent nae maitter whit.

Europe has lairgely been absent frae the heichest levels o that debate, a result o the continent’s langstandin inability tae produce a Silicon Valley-scale titan like Apple, Google or the newcomer OpenAI. While it is a pouerhouse in regulatin, muckle tae the White House’s chagrin, its lack o an Anthropic or Nvidia-sized frontrunner in the AI race has diminished its voice in the safety debate.

Ursula von der Leyen, a pouerfu political figure on the continent as president o the European Commission, the EU’s executive airm, addressed the issue this week by sayin she will invite the main frontier labs for a discussion on hou we can support ongaun industry efforts tae pace the frontier.

Her remarks gained naewhaur as muckle traction as Donald Trump and China’s dismissal o a slowdoon. Without the support o Washington or Beijing, ony sort o pacin is impossible – wi or without Europe’s input.

Lagarde’s speech on Monanday underlined the continent’s relative wakefulness. She asserted that Europe maun develop its ain AI technology and big mair datacentres in order tae nullify the thrait o bein cut aff by the US or China.

Gin Europe invests in its ain AI tech, said Lagarde, the thrait o bein cut aff loses its force. Self-sufficiency wad also gie Europe a chance tae develop its ain answer tae the AI safety crisis.

Margrethe Vestager, wha clashed wi the US tech industry as the EU’s former competition commissioner, co-authored a report this week – A Transformative AI Strategy for Europe – that echoe’d Lagarde’s warnin. AI thraitens tae accelerate a development already affectin Europe, whaur citizens are already bearin the economic and social consequences o decisions taken elsewhaur, she says.

Solutions tae this, accordin tae the report, include biggin mair datacentres – the central nervous systems o AI – and also tae ensure the continent is resilient enough tae withstand AI crises like a pouerfu model slippin oot o control.

Vestager said that Europe does hae a voice in the safety debate through its EU AI Act, the rare piece o AI legislation that sets oot requirements for developers o heich-risk AI systems tae report safety incidents and address risks like a cuttin-edge model evadin control.

But she adds: Europe maun build its ain AI tae strengthen the ecosystem, retain talent, and inspire innovation.

Vestager acknowledges that gin there is an existential thrait, it requires a global solution; one that does nae hand pouer tae the US and China. And that requires a different kind o supranational body.

As for existential thraits tae humanity, we maun first assess the actual level o thrait – then act wi urgency. The UN general assembly is an opportunity tae address this. It is vital tae muve beyond the US-China rivalry narrative. There maun be space for humanity between the egos o leaders like President Xi and President Trump.

Frederike Kaltheuner, an adviser tae the AI Now Institute, a research body, says that Europe is dependent on US technology in ways that will be hard tae undo, frae search engines tae chips tae datacentres. But the narrative that Europe is grievously behind in the race depends on a particular view o artificial intelligence aggressively promotit by American tech bosses.

Because o their doomsday warnins, we are exclusively talkin aboot AI in a very narrow, specific kind o way – namely, the idea that there is sic a thing as the AI frontier, which is gettin better and better, and that access tae the frontier is the maist important thing that Europe needs for its security and economy.

The reality o the market points at somethin else, she says: mony companies are nae adoptin so-called frontier models – like ChatGPT and Anthropic – because o heich costs and fears ower whit micht happen tae their data.

Gin this trend continues, Europe may nae be sae terribly positioned: the EU AI Act has a number o pragmatic measures which address hou consumers micht encounter AI on a day-to-day basis, legislatin ower maitters sic as labellin AI content, and the use o AI in medicine and hirin.

Europe has a competitiveness problem, that’s true, says Itxaso Dominguez de Olazabal, a policy adviser at EDRI, a European digital richts organisation. But US tech boosters are blamin it on laws and protections, and hae promised that deregulation will mak European companies mair competitive … in a race tae the bottom tae see gin we can catch big tech.

Nonetheless, the EU AI Act is nae viewed as model legislation for ither kintras or a vehicle for a global slowdoon. Georgina Kon, a partner at law firm Linklaters, says the act is unlikely tae be adoptit as a global benchmark for several reasons, includin criticisms o the heavy compliance burden, which has led tae delays in pairts o the act becomin active, and a narrow territorial focus on Europe. Those criticisms hae also meant that lawmakin bodies elsewhaur hae looked tae differentiate their approach tae AI legislation frae the EU’s.

The EU AI Act is nae somethin that everybody else in the warld feels constrained by, altho mony tech companies will o course want tae sell tae the EU, she says.

Big tech still has a vast customer base in Europe. Kaltheuner worries that the doomsday narratives will push Europe towards dependency, encouragin it tae adopt US AI models across the economy oot o ill-defined fears ower bein left behind.

There’s a vision o the future whaur Europe’s economy, its public sector, its education system runs on models and on infrastructure that is controlled by a very sma number o companies that happen tae be also in the US. This wad be a sovereignty problem.