A $25 billion Danish pension fund that refused to invest in Elon Musk’s SpaceX is preparing to buy shares in Anthropic when the AI company goes public. The decision highlights a growing debate among institutional investors: how much control should a company’s founders have, especially when billions of dollars are at stake?
Anders Schelde, chief investment officer of Copenhagen-based AkademikerPension, says Anthropic is fundamentally different from SpaceX, despite concerns about the AI company’s own governance structure.
The pension fund has excluded SpaceX from its investment portfolios because of what Schelde considers unacceptable governance risks. But it sees Anthropic as a more attractive option, particularly because of its governance arrangements and stated commitment to AI safety.
The distinction is significant. Both companies are preparing to access public markets, but investors are assessing them not only on their technology and growth prospects, but also on who controls the business and how that power is exercised.
Why AkademikerPension Rejected SpaceX
AkademikerPension’s decision to avoid SpaceX is rooted in corporate governance — the rules and structures that determine how a company is managed and who has the power to make important decisions.
Schelde has argued that SpaceX gives Elon Musk too much control over the company.
The concern is that Musk’s influence could make it difficult for shareholders and independent directors to challenge management decisions, even when those decisions could negatively affect the business.
SpaceX’s proposed share structure gives different voting rights to different classes of shareholders.
Under its IPO prospectus, Musk is expected to control approximately 82.4% of the company’s voting power, despite holding a smaller proportion of its economic ownership.
He would also retain substantial power over the election and removal of directors.
This creates a situation where public investors can own shares in the company but have limited influence over how it is governed.
For a pension fund responsible for managing retirement savings, that is a significant concern.
Why Anthropic Is Different
Anthropic, the company behind the Claude AI assistant, has designed a different governance structure.
It is registered as a Public Benefit Corporation (PBC), meaning it is legally required to consider more than just shareholder financial returns.
Its stated public benefit purpose is the responsible development and maintenance of advanced AI for the long-term benefit of humanity.
Two important structures are central to this arrangement.
1. The Founder LLC
Anthropic’s seven co-founders are organised through an entity called the Founder LLC.
Under the company’s amended IPO filing, this entity will hold a special Class F share carrying 50.1% of the company’s total voting power on matters submitted to shareholders generally.
This gives the founders significant influence over the company’s direction.
2. The Long-Term Benefit Trust
Anthropic has also established an independent body called the Long-Term Benefit Trust (LTBT).
The trust has authority to elect and remove special Class T directors, who together are expected to make up a majority of the company’s board.
Its purpose is to help ensure that Anthropic continues to consider AI safety and its public-benefit mission rather than focusing exclusively on short-term financial returns.
The arrangement is unusual because the trust can influence the board without holding ordinary economic ownership in the company.
Schelde acknowledges that Anthropic’s founder control is not ideal, but considers its structure more acceptable than SpaceX’s.
How Much of Anthropic Does the Pension Fund Want to Buy?
The investment plan is more nuanced than simply committing billions to Anthropic.
According to Bloomberg, AkademikerPension intends to include Anthropic in its passive investment holdings.
Its approach to actively managed investments will depend on the company’s valuation.
Here’s the difference:
Passive investments generally follow a market index or predefined investment strategy rather than relying on a manager to select individual companies.
Active investments involve more direct decisions about whether a particular stock is attractive at its current price.
This distinction matters because Schelde is not saying the fund will buy Anthropic shares at any valuation.
He still has concerns about the financial risks of investing in the AI sector.
Schelde Thinks the AI Market Could End in a Bubble
Despite his willingness to invest in Anthropic, Schelde is not blindly optimistic about the AI boom.
He warned that the sector could eventually experience a bubble, although he believes it is too early to determine exactly when that might happen.
His concern reflects a wider debate among investors.
AI companies are attracting enormous amounts of capital because investors expect artificial intelligence to transform industries, increase productivity and create new sources of revenue.
But the valuations attached to some of these companies are also rising rapidly.
If expectations become too high, companies could struggle to justify their valuations through actual earnings and cash flow.
For pension funds, this creates a difficult balance: avoiding the AI industry could mean missing a major investment opportunity, but investing at excessive prices could put beneficiaries’ money at risk.
Schelde’s position is essentially that investors should participate in the AI opportunity without ignoring the financial and governance risks.
Anthropic’s IPO Has Its Own Governance Critics
Not every major investor agrees with Schelde’s assessment.
New York City Comptroller Mark Levine has urged Anthropic to delay its planned IPO until it addresses concerns about AI safety, corporate governance and the speed of its technological development.
Levine oversees public pension funds with more than $325 billion in assets.
His concern is that Anthropic’s governance arrangement could limit the influence of ordinary shareholders while giving founders and the Long-Term Benefit Trust substantial control.
He has also raised questions about the risks associated with increasingly capable AI systems, including cybersecurity threats and potential misuse.
This creates an interesting disagreement among institutional investors.
Schelde sees Anthropic’s governance model as sufficiently different from SpaceX’s to justify investment.
Levine believes the company’s governance and AI safety questions deserve more attention before it enters public markets.
Neither position should be confused with a guarantee that Anthropic’s structure will work perfectly. Its effectiveness will ultimately depend on how the company and its governing bodies behave when difficult commercial and safety decisions arise.
Anthropic’s Financial Growth Is Impressive, but the Risks Are Significant
The governance debate is happening alongside an enormous financial expansion.
Anthropic’s amended IPO filing, reported by Reuters on November 2, revealed several important figures.
Anthropic’s third-quarter 2026 financial snapshot
$16.96bn
Quarterly revenue
47%
Revenue growth from Q2
$23.29bn
Reported net loss
$895.4bn
Computing spending commitments
Figures from Anthropic’s amended S-1 filing, as reported by Reuters.
There are several details investors should understand.
Revenue is growing, but the growth rate has slowed
Anthropic’s revenue increased from $11.56 billion in the second quarter to $16.96 billion in the third quarter.
That represents 47% quarter-on-quarter growth.
However, the company had recorded 144% growth in the previous quarter.
The business is still expanding rapidly, but the slowdown is something prospective investors will examine carefully.
The company has enormous computing commitments
Anthropic disclosed approximately $895.4 billion in computing-related spending commitments involving infrastructure partners, including Amazon and Google.
Almost half of those commitments are expected before the end of 2030.
These are future spending commitments, not money already paid.
The company also warned that some infrastructure partners depend on third-party financing. If that financing becomes unavailable or more expensive, contracted computing capacity could be delayed, repriced or reduced.
Losses remain substantial
Anthropic’s third-quarter net loss narrowed to $23.29 billion from $66.3 billion in the previous quarter.
It also reported an operating loss of $567 million, compared with operating income of $974 million in the second quarter.
The distinction matters because the net loss includes non-operating items associated with convertible bonds, while operating income reflects the underlying costs and revenues of running the business.
The figures show why investors will need to look beyond Anthropic’s impressive revenue growth when evaluating its IPO.
When Will Anthropic Go Public?
Anthropic has already submitted its registration documents to the US Securities and Exchange Commission.
The company initially announced in June 2026 that it had confidentially submitted a draft registration statement.
Its amended filing was made public in November.
According to Reuters, Anthropic is preparing to begin formally marketing its shares to prospective investors during the week of November 9, 2026.
However, the final share price, number of shares offered and ultimate valuation remain subject to the IPO process.
Some investors have reportedly considered a valuation closer to $1.5 trillion, compared with earlier expectations of around $2 trillion.
That is an indication of how seriously the market is weighing Anthropic’s financial commitments and slowing revenue growth. It is not a confirmed IPO valuation.
What This Means for the Future of AI Investing
The disagreement between AkademikerPension and other institutional investors reveals something important about the next stage of the AI industry.
Investors are no longer evaluating AI companies based only on how powerful their models are or how quickly their customer bases are growing.
They are also examining:
Who controls the company?
Can shareholders hold management accountable?
Are AI safety commitments protected by enforceable structures?
Can the company generate enough revenue to justify its infrastructure spending?
Is the IPO valuation reasonable?
These questions are becoming increasingly important as AI companies move from private funding rounds into public markets.
For pension funds, the stakes are particularly high because they are investing money intended to support people’s retirement.
The ability to identify strong technology businesses is important, but so is avoiding companies whose financial or governance risks are difficult to justify.
Frequently Asked Questions
1. Why did AkademikerPension reject SpaceX?
The Danish pension fund considers SpaceX’s corporate governance structure too risky, particularly because Elon Musk retains overwhelming voting control and substantial influence over the board.
2. Who is Anders Schelde?
Anders Schelde is the chief investment officer of AkademikerPension, a Danish pension fund managing approximately $25 billion in assets.
3. Why is the pension fund interested in Anthropic?
Schelde believes Anthropic has a more acceptable governance structure and a stronger institutional framework for protecting its AI safety mission than SpaceX.
4. Does Anthropic give its founders control?
Yes. Its amended IPO filing provides the seven co-founders’ Founder LLC with 50.1% of total voting power through a special share class. The Long-Term Benefit Trust also has authority over the election and removal of a majority of the board’s special Class T directors.
5. Is Anthropic already publicly traded?
No. It is preparing for an IPO. Its registration process is underway, but the final offering price and listing details remain subject to the process.
6. Is Anthropic a profitable company?
Its third-quarter 2026 filing reported $16.96 billion in revenue and a $23.29 billion net loss. It also recorded an operating loss of $567 million.
Final Thoughts: Investors Are Choosing More Than Just an AI Company
The decision by AkademikerPension is an interesting signal for the technology investment market.
The fund is willing to avoid one of the world’s most valuable technology companies because it considers its governance structure too risky, while preparing to invest in another major AI company with its own unusual founder-control arrangements.
That does not mean Anthropic is without problems.
Its computing commitments are enormous, its financial losses remain significant, and its governance structure is already attracting criticism from other institutional investors.
But the comparison shows that corporate governance can influence investment decisions just as much as technology and growth prospects.
The next big AI investment opportunity may not simply be about which company builds the best AI. It may also be about which company investors believe they can trust with their money.
Would you invest in Anthropic when it goes public, or would its founder-control structure and huge computing commitments make you cautious?
