A Question of Democracy — Part 2: Elon Musk and the Power Beyond the Ballot

Part 2 of our series examining how political systems work, who exercises power and how that power can be challenged. Start with Part 1: Beyond the Labels.

Elon Musk does not have to win an election to become a major presence in other people’s political lives. He can finance campaigning, endorse candidates before an international audience and participate in arguments about institutions whose decisions affect his businesses. Through those businesses, he can also supply services governments need.

None of that establishes that he controls voters or dictates government policy. It does establish a question that democracies should be able to answer: what happens when several different sources of political influence accumulate around one person?

The issue cannot be settled by deciding whether Musk is admirable, objectionable, brilliant or reckless. A political system that relies on the restraint of exceptionally powerful individuals has left an important part of its accountability to personal choice.

The question is how commercial ownership becomes influence over public decisions—and what prevents that influence from escaping effective scrutiny.

1. A right to speak is not an equal capacity to be heard

Musk has the same legitimate interest in political participation as other citizens. Being wealthy should not remove a person’s freedom to criticise a government or support an argument. Successful entrepreneurs may know things that public institutions should hear.

But the existence of that right does not settle the distribution of practical power. A citizen can express an opinion without possessing the resources to put it before millions of people repeatedly, finance a campaign around it or obtain a meeting with those who can implement it.

Our concern is not that some people are persuasive. It is whether money and ownership provide routes to political influence that other citizens cannot realistically access, and whether those routes remain answerable to the public.

Equal voting rights are an essential protection. They do not, by themselves, equalise the conditions under which voters encounter candidates, information and competing explanations.

2. Money shapes the contest before the votes are counted

Campaign resources can pay for staff, advertising, canvassing, research and voter mobilisation. They can sustain a candidate through a difficult period or make an opponent defend themselves against a repeated message. These are mechanisms of influence even when no politician has promised a donor anything.

The US Federal Election Commission explains that super PACs may accept unlimited contributions to finance independent political activity. This is distinct from unrestricted direct contributions to a candidate. Coordination can change the legal treatment, and source prohibitions still matter. “Unlimited” does not mean that every transaction is lawful. [1] [2]

Reuters reported on 12 October 2026 that Musk had spent roughly $366 million during the midterm election cycle to support Republicans, including $276 million since July. That is a dated report of cycle activity, not a final audited total, not all direct candidate donations and not a measure of votes purchased. [3]

The structural issue does not depend on the last decimal place. At that scale, political participation includes the ability to finance substantial campaigning around other people’s choices. Ordinary citizens can combine resources through organisations, but an individual who can deploy enormous sums has a different starting position.

A donor need not issue an instruction for a campaign to value their preferences. Dependence can influence which meetings feel urgent or which positions seem financially sustainable. That is a plausible mechanism to investigate, rather than proof that it operated in every decision.

3. Wisconsin demonstrates the limits of that power

In April 2025, the candidate backed by Musk lost the Wisconsin Supreme Court election despite more than $21 million in spending by Musk and associated groups, according to Reuters. Susan Crawford’s victory preserved the court’s liberal majority. [4]

That result matters. Treating every well-funded intervention as an automatic purchase of political control would make the argument unfalsifiable. If the candidate wins, money supposedly proves control; if the candidate loses, control is merely said to be hidden. Neither approach explains elections.

The defensible conclusion is narrower: concentrated money can change the resources available to a campaign without determining the result. Voters, opposition campaigns, institutions and backlash can limit its effectiveness.

A failed intervention does not remove the inequality in campaigning capacity. Equally, the inequality does not entitle us to describe citizens as incapable of independent judgement.

4. X brings ownership into the political conversation

Musk’s position differs from that of a donor who must purchase space in somebody else’s publication. He owns X, while also using it to advocate political positions. This joins the role of a participant in public discussion to influence over the organisation that hosts and distributes that discussion.

Platform design includes choices about recommendations, moderation, verification, monetisation and access to data. Those choices can affect the environment in which political messages compete. This does not mean the owner personally selects every message that users see.

The accountability question is whether outsiders can examine consequential decisions, identify inconsistent treatment and challenge errors. A general assurance that a platform protects speech does not answer a specific question about how its systems operate.

Nor should criticism of private platform power become an excuse for governments to suppress opposition. State censorship and private control are different mechanisms. Both require scrutiny, with safeguards appropriate to each.

5. What the German research actually found

AlgorithmWatch and the Atlantic Council’s DFRLab studied posts by German politicians and selected anti-AfD organisations ahead of Germany’s 2025 election. They found that the most viral material in their dataset was dominated by references to Musk and his support for the AfD, including highly engaged material involving Alice Weidel. [5]

Their qualifications are important: they did not find clear underlying algorithmic bias in that analysis, and Weidel’s additional attention did not appear to spread more generally to other AfD politicians. The authors recognised that other research designs had produced varying conclusions. Their sample was not a census of everything German voters saw.

That makes the research useful precisely because it separates questions often collapsed together. An owner can become extraordinarily prominent without a study establishing systematic algorithmic favouritism. A message can attract engagement without persuading its audience. An English-language reaction can inflate attention without representing German electoral support.

Reuters separately reported in March 2025 that Musk had promoted right-wing figures in Britain, Italy and Romania. Its analysis of the German election suggested his support had little effect on the AfD’s overall result. We should not substitute his visibility for a demonstrated electoral effect. [6]

6. Influence can work through the agenda

Political influence includes deciding what people feel compelled to discuss. If officials, opponents and journalists spend their time responding to a prominent person’s accusation, that accusation can shape the agenda even when their responses reject it.

This is a mechanism, not a claim that every controversy is deliberately orchestrated. Newsworthiness, outrage and genuine public concern also generate attention. The evidential questions are which subjects displaced others, whose claims received scrutiny, and whether attention changed public understanding or institutional behaviour.

The distinction matters for our own journalism. Repeating an inflammatory claim merely to denounce it can help circulate it. A useful response should establish the claim’s evidence, identify what is missing and return attention to the underlying issue.

The democratic problem is not only unequal access to a microphone. It is unequal capacity to make everyone else answer your questions while their own remain unheard.

7. When the government is also a customer

Government decisions affect businesses through procurement, regulation, taxation and enforcement. These relationships create incentives for business leaders to participate in politics. They also create potential conflicts when those leaders obtain a role in public decision-making.

NASA’s April 2021 announcement awarded SpaceX a $2.89 billion firm-fixed-price, milestone-based contract for its human landing system. That historical award illustrates a substantial public commercial relationship; it is not SpaceX’s current total government revenue or evidence that the award was improperly obtained. [7]

The fact that a firm receives public money is not an indictment. Governments buy expertise and services because they need them. The questions concern the process: competition, technical merit, price, performance, conflicts and the ability to hold suppliers to their obligations.

A useful supplier can become difficult to replace. That raises a further procurement question: has government preserved enough alternative capacity to negotiate, enforce standards and continue operating if the relationship deteriorates?

8. DOGE and the need for independent checks

During Musk’s 2025 DOGE involvement, Reuters reported an expansion of Grok’s use in federal government and concerns about conflicts and sensitive information. Its report also carried a Pentagon response disputing aspects of DOGE involvement in its systems. This was reporting about contested practices and risks, not a court finding of illegality. [8]

Ethics safeguards exist because good intentions do not eliminate financial interests. US Office of Government Ethics guidance describes disqualification and remedies such as exemptions, waivers and divestiture. Whether a particular restriction applies depends on the person’s role, interests and the matter involved. [9]

The questions for public scrutiny are concrete. Who identified relevant interests? Who decided on recusals? Who authorised access to data? Which procurement decisions were insulated from interested parties? What documentation enables an independent reviewer to check the answers?

A system cannot convincingly resolve a conflict merely by accepting an interested person’s assurance that they recognise it. Conversely, identifying a potential conflict does not establish that an official committed a crime. Accountability requires a record and an independent process, not an accusation standing in for both.

9. Infrastructure can carry political consequences

Campaigning attempts to influence decisions. Essential infrastructure can affect what decisions are practically possible. Communications, launch services and other specialised systems may become significant to government operations, with consequences extending beyond the terms of an ordinary consumer purchase.

In July 2025, Reuters reported that Musk had ordered a Starlink coverage restriction during Ukraine’s September 2022 counteroffensive, citing sources familiar with the matter. Its account linked the decision to concerns over escalation and noted uncertainty over subsequent US government exchanges. This should be attributed to the investigation, rather than presented as a complete public record of every instruction. [10]

A further dispute was unfolding as this draft was prepared. Reuters reported on 21 October 2026 that Marco Rubio described Ukraine’s Starlink access as a matter for Musk and commercial contracts, following reporting about service restrictions. Accounts of responsibility and the systems involved were contested. AP’s contemporaneous explanation also reported Musk’s position that the US government controlled Starshield. These reports do not establish a single agreed chain of command. [11] [12]

Starlink and Starshield should not be treated as interchangeable labels, nor should every service restriction be assumed to be a personal political instruction. Contracts, operational arrangements, legal obligations and military risk all matter.

The unresolved institutional question remains significant: who has authority over changes to a service that affects public security, what continuity is guaranteed, and what alternatives exist? Neither a private owner nor a government should be presumed to have unlimited discretion without examining the relevant arrangements.

10. Commercial achievement deserves recognition, not immunity

A serious criticism must acknowledge why governments use these services. NASA’s Inspector General reported estimated savings of between $1.4 billion and $4 billion from SpaceX under the agency’s competitive commercial cargo approach. This is a specific historical estimate, not an endorsement of every SpaceX contract or every commercialisation policy. [13]

Private investment and technical innovation can deliver public benefits. Government ownership alone does not guarantee efficiency, resilience or democratic accountability. Public monopolies can also fail, conceal mistakes and resist scrutiny.

Those points strengthen the case for effective oversight. If a supplier is valuable, the public has an interest in a durable relationship governed by clear obligations. Technical success does not remove the need to examine conflicts, market concentration or reliance on one provider.

Success in one field also does not establish expertise in every other. A claim about migration, policing or constitutional government requires evidence even when its author has built successful technology businesses.

11. Motive is harder to establish than incentive

It is tempting to explain every intervention as a financial calculation. That may sometimes fit, but a political stance can also alienate customers, create opposition or conflict with another part of a business. People can sincerely pursue ideological goals that cost them money.

We should therefore distinguish three things: an observable action, an incentive that could help explain it, and evidence establishing a motive. A company’s interest in a regulatory decision establishes a reason for scrutiny; it does not prove why its leader made a particular public statement.

Musk’s political activity may involve conviction, commercial concerns, personal disputes and ambitions to shape events. Assigning precise weights to those possibilities would require evidence we do not possess. Speculative psychology contributes less than examining decisions and their consequences.

The public does not need a reliable account of someone’s inner life before asking whether their power has adequate safeguards.

12. The strongest defence—and its limits

The defence of Musk’s participation deserves a fair hearing. He is entitled to advocate his views. Voters can reject them. Platforms compete for users. Elected officials remain responsible for government decisions. Private contractors can provide services more effectively than existing public arrangements.

Each point has force. None settles the whole question. The ability to leave a platform does not necessarily give a citizen access to an equivalent audience. The existence of an election does not explain how conflicts were managed between elections. A successful contract does not establish that future dependence is prudent.

The test is not whether there is any accountability at all. It is whether accountability is effective where different forms of power overlap. Can people obtain records, challenge decisions, secure remedies and preserve essential services without depending on the goodwill of the person being challenged?

13. Apply the same standard beyond Musk

Concentrated political influence does not become acceptable when exercised by a donor, publisher or platform owner whose preferences we share. A principle that changes with party allegiance is a preference dressed as a safeguard.

Musk is a revealing case because money, a prominent platform and commercially valuable infrastructure meet around the same individual. He did not invent wealthy political participation, media ownership or conflicts between public office and private interests.

Replacing him with a more agreeable billionaire would not resolve the underlying institutional questions. Nor does calling a country an oligarchy, without specifying mechanisms and limits, do the work of examining it.

The useful inquiry is where concentrated private resources gain leverage over collective decisions, what counterweights exist, and whether those counterweights work in practice.

14. What meaningful accountability would require

No single reform addresses every mechanism. Campaign-finance disclosure helps identify funders, but does not by itself equalise resources. Public funding may broaden participation, but requires fair qualification rules and safeguards against incumbent advantage.

Independent ethics enforcement and documented recusals can protect decisions from financial interests, provided the supervising body has access, independence and consequences it can enforce. Transparency without meaningful review can become a paperwork exercise.

Platform scrutiny needs access to evidence about distribution and moderation, with privacy and security protections. Public oversight must also be constrained so that politicians cannot use it to punish legitimate criticism.

Procurement safeguards include competitive processes where feasible, published performance requirements, audit rights and plans for continuity. Diversifying suppliers can improve resilience, but costs money and may be technically difficult. Those trade-offs should be assessed rather than hidden.

The common requirement is the ability to challenge consequential decisions without first persuading the most powerful participant to permit the challenge.

15. Why this belongs on The Last Days of Planet Earth

Environmental risks, conflict, energy systems, artificial intelligence and communications infrastructure involve decisions whose consequences reach far beyond an election cycle. The distribution of political influence helps determine which risks receive attention, whose costs count and which remedies remain possible.

A society may retain competitive elections while struggling to supervise the private systems on which its public ambitions depend. Recognising that weakness is a reason to strengthen democratic institutions, not abandon them.

Part 1 asked us to look beyond political labels. Part 2 asks us to look at the resources behind political participation. Who can finance the contest, shape the conversation, gain access to officials and supply infrastructure that becomes difficult to replace?

A democracy should not have to choose between benefiting from an entrepreneur’s achievements and scrutinising that entrepreneur’s power.

Musk’s preferences are not a public mandate. His opponents’ dislike of him is not evidence. The standard should be whether public decisions can be examined and challenged through institutions that remain independent of the people whose interests are at stake.

If wealth can repeatedly buy a larger role in shaping collective choices, one vote each is the beginning of the democratic examination—not its conclusion.

References and scope

Research checked 21 October 2026. Historical events are dated above. Reported allegations and disputed operational accounts are attributed; no finding of bribery, unlawful coordination or criminal conflict of interest is asserted. The October 2026 satellite-service dispute requires a fresh check before publication. The analysis of mechanisms and proposed safeguards is editorial argument, not a claim that every mechanism has been proven in Musk’s case.