The UK must follow the US’s surprising step forwards in reining in Big Tech
Rishi Sunak has stated that only governments can tackle the risks posed by AI. So why isn’t he?
As the UK’s much vaunted AI Safety Summit takes place this week, campaigners have provided deserved scrutiny over the hyping of apocalyptic threats over the more pressing issues in AI: algorithmic management at work, surveillance, profiling, and AI’s environmental impact.
The summit itself shows the UK moving from a total laissez-faire approach to an acceptance that some regulation is needed. However, Sunak still claims Britain’s approach is “not to rush to regulate.” But he’s missing a much bigger point. The challenges which new technology poses to our society aren’t a distant nightmare – they’re already here.
The core problem is the power which Big Tech monopolists hold over technologies which are central to our economy and broader society. Far from constraining these modern robber barons, Sunak’s government uses trade policy to cement this power in perpetuity, giving Big Tech giants and any AI technology they develop or acquire broad freedoms from much needed regulation.
Trade rule secrecy
Trade rules are the perfect vehicle for this, given neither parliament nor MPs have any control or oversight over them, much to the chagrin of a growing number of MPs. Just as intellectual property rules in effect handed Big Pharma global monopolies still utilised over medicines back in the 1990s, so today’s e-commerce chapters are handing Big Tech long-lasting control and secrecy over our data and how they use it.
The Big Tech wish list being pushed and already adopted by Sunak’s government has a number of pillars, including trade rules banning governments from asking tech companies operating in their countries to disclose the ‘source code’ behind the algorithmic systems they are using. That means they – and we – are left in the dark to the inner workings of new technologies like AI.
A ban on localised content means governments can’t prevent their citizens’ data being stored outside their jurisdiction, making it harder for them to control how it is used or abused. What’s more, new rules will make it impossible to compel companies to have a local presence in the countries in which they operate, again making it harder to tax and regulate them.
The risks of trading away protections
All of this matters for the regulation of artificial intelligence. AI uses huge data sets to train computers to make decisions, based on instructions from the algorithms in a source code. Trade deals in which countries sign away their ability to look into the under-the-bonnet programming of AI mean they cannot examine its biases or vulnerabilities.
This is particularly dangerous when used in public health systems or for global south governments understanding its impact on racialised populations. For example, a flaw in the source code in medical equipment could have fatal consequences for patients.
As use of AI encompasses ever more business practices, locking away source code from view ensures the stealth of myriad violations of privacy, civil rights, even competition rules. It was only through researchers combing though source code that it was revealed VW was using cars’ onboard computers to cheat emissions tests.
The US’s surprising turn
Last week, the United States showed it has just woken up to the problems when the Biden administration announced it was withdrawing support for proposals it had initially championed at the World Trade Organisation, a corporate charter enshrining the wish list of its huge tech lobby to lock in deregulation and monopoly control of the digital economy.
The Office of the US Trade Representative stated that, as the country was “examining [its] approaches to data and source code, and the impact of trade rules in these areas”, the move was necessary “to provide enough policy space for those debates to unfold.”
While the US has admitted these proposals “might prejudice or hinder those domestic policy considerations”, the UK’s post-Brexit approach has been to pursue new trade agreements that deviate from the EU’s data governance regime, to liberalise data flows and to weaken our ability to regulate AI and Big Tech.
The UK’s post-Brexit approach
The UK’s 2020 deal with Japan, for instance, restricts not only forced disclosure of source code as a condition for exports, but access to algorithms associated with source code. The UK-Australia deal signed in 2021 has similar provisions, as does the Pacific trade deal, known as CPTPP, which the UK acceded to in summer and is expected to ratify in coming months.
While the latter contains some exceptions, critics argue these are far too narrow – and this speaks to a key tension between binding trade agreements and fast evolving AI. Where exceptions are written in, these relate to only the known risks. As the AI Safety Summit is sure to explore, we will see risks and harms popping up as technology develops; but, bound by trade rules, governments may shut off their ability to respond.
Perceiving digital trade, international data flows and AI as dynamic areas of growth, the UK wants to be a prime destination for tech investment. After Jeremy Hunt’s Silicon Valley frolic last month and Kemi Badenoch’s roundtable schmoozing businesses leaders this week, Rishi Sunak is set to meet ultimate tech tycoon Elon Musk for a one-on-one at the summit’s close.
The UK’s Indo-Pacific tilt in trade policy is both a scrambling for new markets outside the EU, but also evidence of a diehard bias toward deregulation and unchecked free trade that is butting up against wider geopolitical alignments, as the US’s announcement shows.
The UK must stand up to Big Tech power
If we truly hope to ensure safety in the digital age, the ability to scrutinise the tools of Big Tech and curb their monopoly power to profiteer from our data is crucial. To achieve this, we need to make sure we have a proper public debate on these issues.
The meeting at Bletchley Park has already shut out much of civil society, but trade deals, even further from democratic accountability, are even worse places to devise public policy.
Sunak’s gambit to position the UK as most attractive data marketplace could merely render us a bleak special economic zone in a race to the bottom: where control over the algorithmic systems penetrating ever more aspects of our daily lives are wrested out of our control by the iron fist of trade deals – and benefit only tech giants who exploit workers, manipulate consumers and squeeze small businesses across the world. It doesn’t have to be this way – the UK must follow the US’s surprising example and stand up to Big Tech’s power.