An entire generation handed over their data without realising it. Every search, every click, every social media platform fed into the advertising models that became the foundations of Google, Meta and Amazon. Europe stood by, regulated too late and let the value slip away. The GDPR arrived when fortunes had already been made. This was no accident. It was an externality that was not corrected in time: companies reaping considerable private gains whilst imposing a cost on society---loss of privacy, concentration of power, undermining of democratic debate--for which they did not pay.
Artificial intelligence is now setting in motion the same mechanism, but with a different resource: human labour. Brett Hemenway Falk, of the University of Pennsylvania, and Gerry Tsoukalas, of Boston University, have set this out in a paper with unsettling findings. Rational firms, fully aware of the danger, are collectively trapping themselves in a race towards automation that destroys the demand on which they depend. Each captures the full cost savings. Each bears only a small fraction of the destruction of demand it causes. The rest falls on its competitors. The final loss is not a transfer from workers to shareholders. It is a deadweight loss that penalises both--just as the plundering of data has ultimately impoverished the media and democratic ecosystems on which the platforms themselves depended.
Six answers, one outcome
Falk and Tsoukalas have scrutinised six policy responses - and this is where the analogy with the data becomes truly uncomfortable. The reskilling of displaced workers partially reduces the externality without ever eliminating it. Employee ownership of the companies that replace them narrows the gap between the private optimum and the collective optimum - but internally shared gains do not restore the demand destroyed among competitors. Negotiation between firms faces a fundamental obstacle: automation is a dominant strategy, and no voluntary agreement is self-enforcing. For years, Europe believed that platforms would self-regulate on data. They did not. The same naivety would be fatal a second time.
Universal basic income, championed by Sam Altman as a response to the age of superintelligence, maintains consumption levels without altering in the slightest the incentive for individual companies to automate further. Worse still: by increasing autonomous demand, it attracts new entrants, further fragments the sectors concerned - and as distortion increases with the number of competing firms, the remedy may exacerbate the problem it claims to cure. The capital tax produces the same lack of result: it affects profit levels, not the automation margin where the externality lies. Five instruments. Five structural failures.
Only a Pigouvian tax--calibrated to the external cost that each firm imposes on its rivals without paying for it--corrects the incentive at source. The revenue from such a tax can fund retraining, gradually reduce the externality and make the tax potentially temporary. The authors even show that better AI exacerbates the problem: each firm gains market share by automating further, but at equilibrium these gains cancel each other out, and only the destruction of demand remains.
The response must be coordinated across Europe
The objection is predictable. A country that taxes automation on its own risks seeing business migrate elsewhere. Fedil, the UEL and their European counterparts will be sure to raise this point--just as tech lobbies have for years raised the spectre of competitiveness to delay any data regulation. The argument of a lack of coordination has long been used as an excuse for inaction. Until collective pressure made inaction more costly than reaching an agreement - as demonstrated by the international minimum tax.
Just as the Prime Minister, 
Europe missed out on the first wave of exploitation through inaction and naivety. Fortunes were made elsewhere; regulation came too late, and the GDPR has neither revived the media outlets that were destroyed nor returned their data to the public. Human labour is the next resource on which considerable private profits will accumulate, with the bill being passed on to the public. Falk and Tsoukalas have at least identified the mechanism. The question that remains is simpler, and more political: will Europe wait, once again, for fortunes to be made? And then have to bear, in some unknown way, the cost of the destruction?
