The First Gilded Age

The Gilded Age¹, a satirical novel co-authored by Mark Twain highlighted the greed, corruption, and speculative frenzy of post–Civil War America. The title referred to a society where a thin layer of gold concealed the very different conditions experienced by those below the top echelon. It symbolised an era that appeared to glitter with opportunity but hid deep social and political decay.

The Gilded Age was an era of rapid technological change. Railroads, the telephone, oil, automobiles and electricity transformed the economic landscape. The US industrialised and, supported by its abundant natural resources and inflows of labour, overtook Britain as the pre-eminent global industrial nation. At the same time, the appearance of near monopolies such as Standard Oil and US Steel was accompanied by the emergence of the ‘robber barons’. Forming trusts to eliminate competition, suppressing labour to force down costs and bribing politicians to create supportive legislation were trademark tactics of the time. This was accompanied by rampant corruption in the political system exemplified by the political patronage of Tammany Hall in New York City and the Credit Mobilier Scandal of the Union Pacific Railroad. Rockefeller, Carnegie, Vanderbilt, J.P. Morgan and Jay Gould were notable figures of the time. Such was their historical impact it is no surprise that they have all been the subject of multiple biographies.

The result of how the economy developed and the actions of the robber barons created the Gilded Age conditions and with it the concentration of wealth. US income inequality reaching the point where the top 1% accounted for nearly 20% of income, and the richest 10% of the population held 70-80% of US wealth. This caused significant social and political unease, eventually provoking a serious backlash and was further fuelled by a stream of revelations concerning corruption of the political process. It culminated in the Sherman Antitrust Act (1890) which prohibited price fixing and collusive practises. This was flowed by the Clayton Act (1914) which restricted anticompetitive mergers and acquisitions and the establishment of the Federal Trade Commission (1914). It was also a period where investigative journalism played a significant role in exposing both the corruption and anti-competitive practices. This was exemplified by the work of Ida Tarbell, whose research and subsequent publications were published in a series of articles and then as a book². Her detailing Standard Oil’s monopolistic practices helped contribute to its ultimate break-up.

To summarise, the Gilded Age was one of rapid industrialisation, societal change, increasing concentrations of power and wealth, market abuse, and corruption with undue political influence being wielded by the corporate ‘titans. The parallels with the modern world are not hard to see.

The Second Gilded Age?

If the preconditions for a gilded age were concentration of power, monopolies, social and political unease, and wealth and income disparities, then they seem to have reappeared. The question of corruption is more difficult in the sense of whether a legal criminal definition has been met, but it seems unarguable that the new economy and government are very tightly linked with each seeking to influence the actions of the other.

Some economists have recently highlighted the extreme nature of wealth concentration in the US. For example, some estimates suggest that roughly 19 households in the US currently control wealth equivalent to 12% of US national income. The previous peak during the Gilded Age was approximately 4%. The comparison is startling.

Wealth Concentration in the US: two Gilded Ages?

AI and the New Means of Production: The Modern Titans

In the industrial age, the titans controlled either the ‘means of production’ or the financial institutions that funded them. Financial institutions remain important today, but arguably in today’s world where intangible products have come to the fore the means of production have changed significantly. Production now comprises three elements: compute, algorithms and data. Algorithms and data are inseparable since AI depends upon the interaction of the two. Currently datacentres are the critical physical infrastructure and an integral part of the new ‘means of production’. Whilst they will remain critical infrastructure, as capacity builds and supply and demand come more into balance, they may assume more utility-like characteristics, as power generation did with the development of the electricity supply industry as the world industrialised.

However the structure develops, it is clear that the LLM segments of the AI industry will be dominated by the relatively small number of players able to achieve and maintain the necessary scale. They will be critical cogs in the economy but their role and influence will extend well beyond this. As discussed in the Anthropic/Department of Defence ‘debate’, the impact of AI on areas such as privacy and defence will be profound. Moreover, the interaction of the ‘tech bros’ with politics and government is already obvious, not least in the performative spectacle of Elon Musk brandishing a chainsaw at the 2025 Conservative Political Action Conference. More broadly, it is natural that industry leaders will seek to influence policy to the benefit of their organisations particularly in regards of access to data, the raw material of AI.

Regulation, Power and Public Backlash

There is no doubt that the impact of AI will be long-lasting, bringing with it the potential for huge positive rewards for society. Industrialisation brought similar benefits, but it also created serious societal stresses that ultimately demanded a legislative response. Health, education and workplace safety all required government intervention. Inevitably there were time lags, not least because the new titans of industry resisted legal remedies, arguing that markets and self-regulation would be sufficient. Naturally they viewed any form of legislation as a potential drag on profits. This occurred irrespective of market structure, but history shows that where monopolies exist, they find it very hard to resist abusing their market power. They also typically enter the political arena to protect these monopoly powers. On top of this, there are the dangers that arise when politics, government and business become too closely intertwined - where legal lines can be crossed. Corruption was a feature of the Gilded Age à la Tammany Hall and the Credit Mobilier scandal, but criminal corruption is not necessary for there to be concerns, it is entirely possible that undue influence can be exerted with no criminal element.

What comes next?

As the First Gilded Age demonstrated, widening disparities in living conditions, income and wealth bred resentment and political instability. Those below the elite layer were galvanised by recurrent revelations concerning the behaviour of their political and industrial leaders, hence building support for legislative intervention.

Given the similarities with today, it is not difficult to envisage a similar scenario once again unfolding. If the historic patterns are relevant, it is likely that some form of exposure of wrongdoing will be required to create the necessary momentum. It could, for example, follow from the efforts of investigative journalism akin to Ida Tarbell’s exposure of Standard Oil. Equally it could a modern-day version of the industrial accidents and steam boiler related injuries/fatalities to overcome the resistance of powerful vested interests.

Consumer-oriented legislation

There is already a growing body of evidence of the harm that can follow from the absence of appropriate controls in generative AI.

For example, we know that:

The main line of defence is often clothed in more sophisticated language but, in essence it portrays social media and AI as conduits rather than creators. Responsibility therefore lies not with the tools but with those who use them. The conduit is the ‘mailman’ not the ‘publisher’.

A recent example came from Elon Musk during the UK controversy surrounding the use of GrokAI to generate and edit sexually explicit and non-consensual images of women and children on X³. The initial defence was to deny any knowledge of such images. This was followed by the argument that blaming Grok was akin to ‘blaming the pencil for what the artist chooses to draw’. Musk also suggested the more sinister motive – the suppression of free speech - lay behind the complaints.

One would expect examples such as these to continue increasing the pressure for legislation. There are two obvious routes. The first is to treat the service as a publisher, making it responsible for content posted or generated though its platforms. There are legitimate arguments on both sides, many of which have been rehearsed in other contexts ranging from tobacco through to gun control. The example of firearms demonstrates that the assignment of liability to the creator of a tool rather than the user rarely gains traction. While there are some examples such liability accruing where alcohol is served to clearly intoxicated customers, the examples are limited. The second route is to require users to have verified identities, ensuring that should the law be transgressed the individuals responsible can be held accountable. This would be the equivalent of creating a digital identity card for on-line content creators. In the current environment this may appear draconian and politically unacceptable as the inability to entrench meaningful gun registration requirements demonstrates. However, public sentiment can change, and self-regulation can be replaced by legislation if self-regulation is perceived to be acting in the interests of the companies rather than the customers.

Whilst there are important implications for society, for investors it is worth bearing in mind that the current legislative framework might be considered the most benign possible. Should this change the there will be a direct impact on profitability.

‘Macro’ oriented legislation

If legislation is forthcoming to manage the new age it is unlikely to be limited to the direct operation of AI alone. The concentration of wealth and control of digital media channels directly impinge upon the democratic process. The tech industry has acted in a manner no different from its predecessors by seeking to influence policy to ensure the most favourable conditions for maximising profits. The potential difference today is the extreme concentration of power and wealth, which has not just reached but significantly exceeded that of the original Gilded Age. In this they have been aided by characteristics of their businesses. As producers and conveyors of global intangible products, they have been able to adopt tax strategies to minimise their effective tax charge. The ability to place intellectual property in a particular jurisdiction opens up the opportunity to take advantage of taxation differentials in a manner that is less available to companies with physical capital equipment. Few have spurned this opportunity. The ability to choose personal tax domicile may also give rise to public concerns. It is everyone’s right to arrange their tax affairs in a manner most efficient for them. Hence re-domiciling to tax havens such as Monaco or from higher to lower tax states in the US.

In a world governed by national tax legislation and complex tax legislation, attempts to address tax arbitrage have proven problematic. National (and intra state) interests naturally clash and some tax jurisdictions continue to behave as ‘free-riders’, attracting additional revenues by offering low-tax regimes. One would expect the question of tax domicile to remain a hot issue in a world burdened by fiscal pressures.

Finally, the growing concentration of wealth and power is likely to become an increasingly important political and societal concern. There will be repeated efforts to address the challenges it creates through both the taxation system and greater scrutiny of monopolistic practices, in a manner reminiscent of the anti-trust movement that emerged during the First Gilded Age.

1 Twain, M. and Warner, C.D. (1873) The Gilded Age: A Tale of Today.

2 Tarbell, I.M. (1904) History of the Standard Oil Company. New York. McClure, Phillips & Co.

3 Early Day Motion 2588, January 2026, UK Parliament.