• Binance Application: The world's largest cryptocurrency exchange formally applied for a pan-European MiCA license in Greece through the Hellenic Capital Market Commission.
• MiCA Regulation: The European Union introduced MiCA in 2023 to standardize regulatory and compliance requirements for crypto firms operating across EU jurisdictions.
• License Deadline: Digital asset firms must obtain MiCA licenses by July 1, or face potential shutdown of operations in various EU countries.
• Greece Selection: Binance chose to file in Greece rather than Malta or Latvia, and established a holding company there with indefinite duration indicating long-term strategic commitment.
• Fast-Track Process: The Hellenic Capital Market Commission engaged major accounting firms including Ernst & Young and KPMG to review the application.
• European Operations: Binance currently operates in at least six European countries with over 20 million regional customers and offices in multiple nations including France.
• Regulatory History: The company faced compliance scrutiny from French regulators and underwent onsite financial inspections; it settled a criminal probe in the United States in 2022.
• Approval Outlook: Industry sources indicate companies typically apply for MiCA licenses only after months of regulator engagement and reasonable confidence of approval.

Binance, the world’s biggest cryptocurrency exchange, has formally applied for a pan-European license known as MiCA (Markets in Crypto-Assets) that digital asset firms operating in the continent must obtain before July 1. The European Union introduced the MiCA regime in 2023 as a way to standardize the regulatory and compliance regime for crypto companies, while also streamlining the process for those firms to set up shop across different EU jurisdictions.

Recommended Video

Despite earlier speculation that Binance would seek its license in Malta or Latvia, the exchange opted to file in Greece, where it has also set up a holding company. The Binance application was first reported by the country’s English-language news site D News. The site reported that the process is on a fast-track under the Hellenic Capital Market Commission (HCMC), and that the regulator has asked major accounting firms, including Ernst & Young and KPMG, to help it review the proposal.

A Binance spokesperson confirmed in a statement that the company had sought a MiCA license in Athens, and that it was in active discussions with the HCMC. They added that Binance views the MiCA regime as offering regulatory clarity and a clear framework for innovation.

Greece is perhaps a surprising choice for the application given that Athens is not known as a major financial center, and its regulators are not known as overtly crypto-friendly as is the case with Malta. If the application is approved, it would likely result in Binance creating a significant footprint in the country in the form of a major corporate office and a sizable number of staff.

“Greece is an important contributor to the EU’s economic framework, with an economy growing above the EU average and a strong regulatory environment that promotes financial stability, transparency, and investor protection,” said the Binance spokesperson.

The HCMC did not reply to a request to comment on the application or its timeline. According to D News, Binance’s Greek holding company “has been incorporated for an indefinite duration, suggesting a long-term strategic presence rather than a temporary arrangement.”

Key deadline looms

When the European Union introduced MiCA, many crypto observers hailed it as a first-of-its kind regime that effectively combined consumer protection and regulatory clarity. At the time MiCA went into force in 2023, EU member countries gave firms with existing crypto operations until the end of June to obtain a license. Firms that fail to obtain one in time could be forced to shut down operations in various countries.

This deadline puts pressure on Binance, which currently operates in at least six European countries through licenses issued by various national regulators, and counts over 20 million customers in the region. The company has offices in several European countries, including France, where it has repeatedly tangled with regulators who have accused Binance of lax compliance policies. In October, Binance, which in 2022 agreed to settle a major criminal probe in the United States, confirmed that it has been subject to onsite inspections by French financial regulators.

In recent months, a handful of countries led by France have expressed unease over the MiCA regime, which lets firms with a license “passport” into every EU country, and pushed to hand crypto regulation over to a central authority, the European Securities and Markets Authority (ESMA). Crypto industry lawyers and others, however, have warned it would be disruptive to upend the MiCA regime at this stage. It’s unclear for now if the ESMA proposal has meaningful traction.

Barring a major change to the MiCA regime, Binance is likely well-positioned to receive a license from the Greek regulators. According to a person familiar with the process, companies typically engage with regulators for months before applying for a MiCA license, and only do so when they have a reasonable confidence it will be approved. When a license is granted, crypto companies can operate and engage in marketing across the EU, though they are still subject to certain national laws, including those related to consumer protection.

Join us at the Fortune Workplace Innovation Summit May 19–20, 2026, in Atlanta. The next era of workplace innovation is here—and the old playbook is being rewritten. At this exclusive, high-energy event, the world’s most innovative leaders will convene to explore how AI, humanity, and strategy converge to redefine, again, the future of work. Register now.

About the Author

By Jeff John RobertsEditor, Finance and Crypto

LinkedIn iconTwitter icon

Jeff John Roberts is the Finance and Crypto editor at Fortune, overseeing coverage of the blockchain and how technology is changing finance.

See full bioRight Arrow Button Icon

• Protocol Release: AdGuard VPN has open-sourced TrustTunnel, its VPN protocol, making it available for public exploration, auditing, and use in other projects.
• Core Technology: TrustTunnel is a modern, secure, mobile-optimized VPN protocol that powers AdGuard VPN across mobile, desktop, and browser extension platforms.
• Design Rationale: The protocol was created to address weaknesses in existing solutions like OpenVPN, WireGuard, and IPSec, which are susceptible to detection and blocking in restricted internet regions.
• Technical Approach: TrustTunnel uses TLS-based encryption and HTTP/2 or HTTP/3 transport to blend with standard HTTPS traffic, making it resistant to throttling and deep-packet inspection while maintaining speed.
• Architecture Features: Each connection operates on dedicated streams combining packets for efficient transmission and includes optimization for mobile platforms and unstable network conditions.
• Transparency Goal: Open-sourcing allows AdGuard VPN users to audit the protocol directly and aligns with the company's support for open-source software principles.
• Implementation Resources: The release includes protocol specifications, complete reference implementation code under permissive licensing, and command-line clients for Linux, Windows, and macOS plus iOS and Android applications.
• User Capabilities: TrustTunnel clients support flexible routing rules, fine-grained traffic control by domain or application, and real-time request logging for complete connection transparency.

Today is a big day for us, and for everyone who cares about transparency, privacy, and having full control over their own traffic. We’re finally open-sourcing the protocol that powers AdGuard VPN. And it now has a name: TrustTunnel.

For a long time, we’ve wanted to make the protocol public. Many of you asked for it, and we always said: yes, we will, it’s only a matter of time. Well, the time has come.

🎉

TrustTunnel is now open-source, free to explore, audit, build upon, and use in your own projects.

What is TrustTunnel?

At its core, TrustTunnel is a modern, secure, mobile-optimized VPN protocol. It’s the very same technology that has been running inside all AdGuard VPN apps: on mobile, desktop, and browser extensions.

Why TrustTunnel? Because we needed something better

There are plenty of VPN protocols out there, so why create our own, some might ask. That is because we’ve seen in practice the faults of popular VPN protocols, especially in countries with tight restrictions on internet access. Protocols like OpenVPN, WireGuard, and IPSec share common weaknesses: they are easy to detect and block at the network level, and attempts to conceal VPN traffic often reduce speed. Traditional approaches “wrap” VPN data in a TCP connection and mimic normal web traffic, but TCP’s way of confirming every piece of data creates delays and makes the connection slower.

Unlike those conventional VPN protocols, TrustTunnel is engineered to blend in with regular HTTPS traffic, making it far harder to throttle or block and helping it slip past deep-packet inspection, all while preserving strong privacy and security. It achieves this through TLS-based encryption, the same standard that secures HTTPS, and by leveraging HTTP/2 or HTTP/3 transport, which are ubiquitous on the web. Each connection runs on its own dedicated stream, which combines packets for faster, more efficient transmission. It is also optimized for mobile platforms and performs well even in unstable network conditions.

A protocol you can use, run, tweak, extend, and build upon

By releasing TrustTunnel, we hope to achieve two things. First of all, we want to finally show our users what protocol is powering AdGuard VPN, thus allowing them to audit it openly. At AdGuard, we have always been staunch supporters of the idea of open-source software, and many of our products have long been open source. AdGuard VPN was lagging behind in this regard, but with TrustTunnel being released publicly, it is starting to catch up.

But most importantly, we want to change the status quo in the world of VPN protocols and offer an alternative to existing solutions. That said, we do not want it to be just a PR stunt, when the protocol’s code is de-facto ‘open source,’ but only one VPN service actually runs it. We believe in free and open-source software (FOSS) and want TrustTunnel to be used widely, including by other VPN services. We believe this is the right way to go about open source development, and we hope the community will participate in the TrustTunnel evolution. We welcome any contribution, whether it is a feature request, a bug report, or even a direct contribution to the app’s development.

What have we done to make this possible?

1. We are publishing the first version of the TrustTunnel specification.
2. We are releasing the complete code of our reference implementation of the TrustTunnel server and its clients under a very permissive license.

You don’t have to install AdGuard VPN to use TrustTunnel. You can configure your own server and use open source TrustTunnel clients:

• Command-line TrustTunnel clients support Linux, Windows, and macOS

• We are also releasing two client apps for iOS and Android

TrustTunnel clients already have a lot of functionality, they allow you to:

• Use flexible routing rules to decide which requests go through the tunnel and which stay on the local network

• Exercise fine-grained control, separating work and personal traffic, routing specific domains or apps, and tuning network behavior without complicated setup

• Benefit from a real-time request log that provides full transparency into where the device sends traffic, how routing rules apply, and which connections use the tunnel

Useful links

This is a long-awaited moment for us. We promised to open-source our protocol, and today we’re delivering on that promise. With TrustTunnel now open source, users and developers alike can explore, self-host, and build on the technology.

To get started, check out the following resources:
TrustTunnel website
TrustTunnel open-source repository on GitHub
TrustTunnel app for iOS
TrustTunnel app for Android

• AI-Generated Disinformation Threat: Artificial intelligence can now produce fake videos, audio, documents, and false official results that are highly convincing and nearly indistinguishable from authentic content.
• Election Integrity Redefined: Election security now requires protecting not just voting systems and ballot counting, but also public perception and trust in the electoral process itself.
• Psychological Impact of False Content: Fabricated election-related content does not need to be true to cause damage; it only needs to be believed long enough to suppress voter turnout or undermine confidence.
• Rapid-Response Communication Plans: States and localities must establish immediate communication strategies to counter false information through the same social media channels where misinformation spreads.
• Information Verification Infrastructure: Election offices should designate public websites and social media accounts as authoritative, truth-centric platforms with pre-written responses to potential false claims.
• Transparency as Trust-Building: Public audits, observable safeguards, logic-and-accuracy tests, and visible verification processes help build voter confidence in election integrity.
• Voter Education as Defense: Citizens who understand how elections actually function are more resistant to manipulation and less likely to accept false claims about election processes or results.
• Bipartisan Vulnerability: False content originates from multiple political perspectives and sources, making election confidence a shared interest across all political affiliations.

X

Story Stream

recent articles

• Video: Global Warming Lorem Ipsum Dolor Sit ...
• Article: Global Warming Lorem Ipsum Dolor Sit ...
• Article: Global Warming Lorem Ipsum Dolor Sit ...
• Entry: Global Warming Lorem Ipsum Dolor Sit ...
• Video: Global Warming Lorem Ipsum Dolor Sit ...

If Americans thought the 2016 and 2020 elections were messy, just wait until Mr. Artificial Intelligence shows up for the 2026 midterms with his own election disruptions in hand.

If you are an election official, an election integrity advocate, or even a candidate, you should start creating an anti-AI election disinformation plan – yesterday.

Protecting elections has historically meant securing ballot boxes, counting rooms, and voting machines. In 2026, that will still matter, but will no longer be enough. The greatest threat now to confidence in elections may not be whether votes are counted accurately, but whether voters believe anything they see, hear, or read about the process.

We are entering the first election cycle in which artificial intelligence can produce fake videos, audio, documents, and bogus “official” results virtually indistinguishable from the real thing. Not crude hoaxes. Not obvious scams. But highly convincing, emotionally charged, algorithmically targeted fabrications designed to spread faster than the truth can catch up.

AI changes the nature of election security.

The integrity of an election is not just whether the systems work, but whether the public trusts that they work. Imagine a realistic “official” election day video on social media announcing that polls are closed due to a “security incident.” Or a convincing audio clip of a county supervisor claiming certain ballots won’t be counted. Perhaps a professional-looking website posting fake precinct results that discourage voting.

None of these needs to be true to do damage. They only need to be believed – long enough to suppress turnout, inflame suspicion, or permanently undermine confidence.

This is not science fiction. The tools to do this already exist; we see them every day, and they are getting easier to use, cheaper to access, and harder to detect.

For years, election integrity advocates have focused – correctly – on hard security: paper ballots, audits, chain of custody, cyber defenses, and physical safeguards. These are essential and should continue to be strengthened. But AI-driven misinformation creates a different kind of vulnerability - an attack on perception.

And as the saying goes, “perception is the reality.” You can run a perfectly secure election and still lose public trust if millions of voters are flooded with believable lies they see with their eyes about what is happening.

That is the new election integrity battlefield.

The question “how do we know what’s true?” is not theoretical but operational. In 2026, it will be unavoidable.

This means election integrity planning must expand beyond machines and procedures to withstand, adapt, and recover from false, misleading, or manipulative information.

First, states and localities must create election-related rapid-response communication plans. When false information spreads, official sources must respond immediately, clearly, and through the same channels the lie is traveling – likely Facebook, X, and other social media. Silence creates a vacuum where doubt seeps in like water.

Election offices also must treat public communication with intentionality as a core, primary security function. Websites must be truth-centric platforms where accurate information is posted immediately. In addition, social media accounts, clear chains of public messaging, and pre-written contingency explanations should be part of every election security plan. Voters should know in advance where to find authoritative information and how to recognize it.

As with everything related to election integrity, transparency is king. Routine audits (real ones, not useless Risk Limiting Audits), public logic-and-accuracy tests, clear reporting processes, and observable safeguards don’t just improve security – they build confidence. In a world where anything can be faked, trust is built by systems that are open, visible, and verifiable.

Election officials and election integrity advocates need to be realistic about an effective response. No AI detector will be perfect. No platform will stop all false content. That means the most important defense is not just technical – it is civic. A public that understands how elections actually work is harder to deceive.

If voters know when results will be available, fake “leaked outcomes” lose power. If they know how ballots are secured and counted, viral rumors about manipulation become easier to dismiss. Voter understanding is not just education – it is key infrastructure.

None of this is partisan, and no party can claim the moral high ground. Fake memes and graphics leak out of liberals and conservatives, Democrats and Republicans, not to mention crazy bots from heaven-knows-where. When confidence in elections collapses, it doesn’t matter who wins. The system itself is perceived as the opposition.

In the age of AI, the great vulnerability to our elections is not just voting machines, inaccurate voter rolls, and the increasing threat of noncitizen voting. We must add citizens who accept lies as truth, either because they don’t know any better, or worse, because they choose to.

We cannot secure elections in 2026 only with the tools of the past. The threat matrix has changed. And election integrity must change with it.

As election integrity advocates in the age of AI, the ultimate election security system is not just what we build, but what we believe. If we fail to secure both, we will lose more than elections – we will lose the consent of the governed.

Election offices and election integrity advocates must take note and get crackin’ – midterms are right around the corner, and the most important piece of election equipment might not be the public’s ability to tell the difference between truth and a very good lie.

Kerri Tolockzo is Director of Public Affairs for Proven Media Solutions and the former Executive Director of Election Integrity Network.

• AI Album Launch: ElevenLabs released “The Eleven Album,” citing co-creation with artists such as Liza Minnelli and Art Garfunkel and making the project available via elevenlabs.io and Spotify.
• Eleven Music System: The company’s Eleven Music model can generate full compositions from prompts, offer granular edits to lyrics, timing, and instrumentation, and export up to six studio-quality stems.
• Creator Rights: Artists reportedly retain full authorship and commercial rights while streaming revenue flows through their own channels, aligning with ElevenLabs’ stated commitment to creator control.
• Genre Span: The album covers rap, pop, R&B, and EDM, featuring contributions from Liza Minnelli, Art Garfunkel, Patrick Patrikios, Willonius, IAMSU!, Demitri Leiros, Emily Falvey, Sunsetto, Kondzilla, Chris Lyons, Michael Feinstein, and AI acts Kai and Angelbaby.
• Minelli Track: “Kids Wait Till You Hear This” is an EDM piece using Minnelli’s actual voice as she repeats title phrases, and she praised the project for respecting artists’ voices, choices, and ownership.
• Garfunkel Track: “Authorship” pairs spoken-word lyrics with piano and rain sounds, with Garfunkel noting the tech honors musicianship and keeps humans central.
• Iconic Voice Marketplace: Minnelli and Garfunkel join ElevenLabs’ marketplace for celebrity AI voices, while ElevenLabs also struck deals with Kobalt Music and Merlin to involve artists and songwriters in model development and revenue.
• Business Moves: CEO Mati Staniszewski highlighted the album as a milestone in artist-AI collaboration, the company partnered with Matthew McConaughey for a Spanish audio newsletter, and it offered a $100 million tender valuing ElevenLabs at $6.6 billion.

Several well-known music artists have turned to ElevenLabs‘ AI as a virtual muse — and record producer.

ElevenLabs on Wednesday launched “The Eleven Album,” which the artificial-intelligence audio research and product company touted as the result of an unprecedented partnership with flesh-and-blood artists — like Liza Minelli and Art Garfunkel — to “co-create” original music with its AI-based model for generating “studio-quality compositions.” The album is available to stream at elevenlabs.io/eleven-album and on Spotify.

Related Stories

[

](https://variety.com/2026/film/news/return-to-silent-hill-alternate-reality-game-1236632804/)

Cineverse Launches 'Return to Silent Hill' Alternate Reality Game Ahead of Film's Release (EXCLUSIVE)

According to the company, its Eleven Music system can generate full compositions from simple prompts and can help artists “spark new ideas and accelerate their creative workflows.” The Eleven Music model also can let musicians edit tracks with “granular control,” adjusting lyrics, timing and instrumentation, as well as download up to six studio-quality stems for additional mixing and arrangement.

Popular on Variety

Eleven Music lets artists “use AI to expand their creative range while maintaining full authorship and commercial rights,” according to ElevenLabs. The artists who participated in “The Eleven Album” project released their tracks through their own streaming channels, which ensures “all streaming revenue goes back to the artists, reflecting ElevenLabs’ commitment to creator control, consent and fair compensation.”

“The Eleven Album” project spans multiple genres, including rap, pop, R&B and EDM. The album includes original tracks from Liza Minnelli; Art Garfunkel; Patrick Patrikios; Willonius; IAMSU!; Demitri Leiros; Emily Falvey; Sunsetto; Kondzilla; Chris Lyons; Michael Feinstein; and AI music artists Kai and Angelbaby.

Minelli’s track, “Kids Wait Till You Hear This,” is an EDM track that features her repeating the title and the phrase, “All eyes on you,” ending with: “All eyes on me.” (It’s her actual voice — not an AI re-creation, according to her rep.)

Minnelli, 79, in a statement provided by ElevenLabs, said, “I’ve always believed that music is about connection and emotional truth. What interested me here was the idea of using my voice and new tools in service of expression, not instead of it. This project respects the artist’s voice, the artist’s choices, and the artist’s ownership. I grew up watching my parents create wonderful dreams that were owned by other people. ElevenLabs makes it possible for anyone to be a creator and owner. That matters.”

Meanwhile, Garfunkel’s “Authorship” is a spoken-word track, backed by a gentle piano melody and the sound of falling rain “Music has always evolved alongside technology, from microphones to multitrack recording,” Garfunkel, the 84-year-old singer-songwriter who was once a close collaborator with Paul Simon, said in a statement. “What impressed me about this experience was the respect for musicianship. The human remains at the center. My voice plus the technology simply opens another door.”

Both Minelli and Garfunkel are participants in ElevenLabs’ Iconic Voice Marketplace, which lets companies request access to AI-generated voices of celebrities for campaigns, partnerships and creative initiatives.

ElevenLabs previously announced deals with Kobalt Music and Merlin to allow artists and songwriters they represent to participate in the development of Eleven Music models and AI music revenue streams.

“At ElevenLabs, our mission is to reimagine how humans interact and create with technology — and ‘The Eleven Album’ marks a major milestone: the first large-scale artist-AI collaboration bringing together talent representing over 1 billion streams, multiple Grammy Awards, and decades of influence across modern music,” ElevenLabs’ CEO Mati Staniszewski said in a statement. “We’re incredibly grateful to every artist who chose to trust us and explore what becomes possible when AI amplifies human creativity.”

ElevenLabs recently announced a partnership with Matthew McConaughey, who also invested an undisclosed amount in the company, to use ElevenLabs’ technology to create a Spanish-language audio version of his newsletter “Lyrics of Livin’” in his own voice.

Earlier this year the company announced a $100 million tender offer to let its employees cash out their stock, a transaction that valued the company at $6.6 billion, double its Series C funding-round valuation at the beginning of 2025.

Jump to Comments

A few days before Christmas, the Delaware Supreme Court overruled the state’s Court of Chancery and restored to Elon Musk the $56 billion in stock options he earned at Tesla by increasing the value of the company 12-fold between 2018 and 2024.

That’s good news—not only for Musk but also for everyone interested in seeing Delaware continue as the world’s preeminent corporate-law jurisdiction. No court system that invalidates a corporate transaction approved by a large majority of a company’s shareholders could seriously claim to provide fair and efficient corporate laws administered by economically sophisticated judges. (To reach this result, the Court of Chancery had said that the disclosure the shareholders received was inadequate because, though it clearly explained all the economic terms of the package, it did not explain how flawed—in the court’s view—were the procedures used by the directors in approving it.)

However, the good news stops there. Once you get past the headline holding, everything else about the court’s decision bodes ill for Delaware.

In 2018, when the company’s market capitalization was about $45 billion, Tesla’s shareholders approved an extraordinary compensation plan for Musk. Under the terms of the arrangement, Musk could receive up to 12 payments, each consisting of options to purchase 1 percent of the company’s then-outstanding common stock. The value of the options could, and eventually did, aggregate to about $56 billion—by far the largest executive compensation package in history. But, to earn each payment, Musk had to boost Tesla’s market capitalization by an additional $50 billion, as well as meet another target related to the company’s revenues or earnings.

Thus, to get all 12 payments Musk had to increase the value of the company more than 12 times over, to $645 billion. In other words, to earn the full $56 billion, Musk had to produce $600 billion in value for the shareholders. Amazingly, Musk delivered, and today, Tesla’s market capitalization stands at about $1.4 trillion.

Soon after the package was approved in 2018, a lone Tesla shareholder sued, claiming that the deal was unfair to the company. In 2024, the Court of Chancery agreed and rescinded the transaction, leaving Musk with nothing.

To reach this result, the court applied the so-called transaction-specific controller doctrine. Under traditional principles of Delaware law, controlling shareholders are held to high fiduciary standards. But to be a controlling shareholder, a person had to have either majority voting control or the functional equivalent of such power, meaning he could broadly dictate corporate policy to the board of directors.

Critically, under these traditional principles, you can know with reasonable certainty before embarking on a transaction whether you’re a controlling shareholder. Your status as a controlling shareholder is not dependent on what you may or may not do in the future in connection with the potential transaction.

In a string of cases over the past 25 years, however, Delaware’s Court of Chancery developed a new doctrine: if a person exercises too much influence over any particular decision the board makes, then that person may be considered a transaction-specific controlling shareholder. This is true regardless of how many shares he owns or whether he can control corporate policy generally.

This transaction-specific controller doctrine, which had no basis in prior decisions of the Delaware Supreme Court, can be applied only after a transaction is negotiated and approved. Thus, a person can never know ahead of time whether he is a transaction-specific controller.

This was the doctrine the Court of Chancery applied in the Musk case. In 2018, Musk owned only about 22 percent of the Tesla shares. He had neither majority voting control nor the equivalent of such control. The court found, however, that Musk exercised too much influence over the negotiation process for the compensation package—because, for example, negotiations temporarily stopped when Musk was busy and recommenced when he was ready to continue, and because the board accepted the numbers Musk suggested. There was some truth to this: the board did behave as the court said. The reason, of course, was not that Musk had the power to control the board. Rather, he had the power to walk away and go to Mars with SpaceX. He had a bargaining power, not the power of a controlling shareholder, whether on a transaction-specific basis or otherwise.

Because of the obvious problems with the transaction-specific controller doctrine, the Delaware General Assembly abolished it legislatively last summer. The new legislation does not apply to pending cases, however. So in hearing the appeal in the Musk case, the Delaware Supreme Court had to decide what to do about the Chancery’s problematic doctrine.

The simplest way to restore the market’s trust in Delaware courts would have been for the state’s supreme court to abolish the doctrine, ruling that it runs counter to the general principles of Delaware corporate law. But the court’s ruling never even mentions the doctrine. Rather, noting that the various justices had different opinions regarding the Chancery’s claims that Musk and the other directors breached their fiduciary duties, the Delaware Supreme Court left this holding in place—thus implicitly affirming the transaction-specific controller doctrine.

But if the directors breached their duties in approving the package, how could the Delaware Supreme Court restore Musk’s $56 billion compensation package?

The answer is that the court only reversed the remedy the Court of Chancery applied—that is, rescission of the compensation package. This part of the court’s reasoning, at least, was correct. Under clear principles of Delaware law, a court may rescind a transaction only if rescission returns both parties to the positions they held prior to the transaction. While rescinding the package may perhaps have returned Tesla to its status quo ante, there was no way this could be done for Musk. He can never get back the time and effort he spent working for Telsa.

But if rescission was not the right remedy, what was? Presumably, Musk should have been forced to return the difference between the value of what he actually received and whatever (in the court’s view) was the fair value of his services. It is up to the plaintiff, however, to prove that he is entitled to a remedy, and since the plaintiff’s lawyers requested no remedy except rescission, there was no basis for computing damages; so the Delaware Supreme Court said that it would award nominal damages in the amount of $1.

At this point, things get very strange. Having blamed the plaintiff’s lawyers for failing to request the correct measure of damages, the court then declined to remand the case to the Court of Chancery to determine the appropriate fee for those lawyers, which would be the usual thing to do. Instead, the court decided on its own that those lawyers are entitled to a fee based on the legal doctrine of quantum meruit—that is, a fee based on how much they deserve based on benefit the case has conferred on the company.

Since the company received only $1 in damages (and no doubt spent tens of millions litigating the suit), you might think that the fee for the lawyers would be, at most, some fraction of a dollar. But no: the plaintiff’s lawyers will get $54 million—four times the value of their billable hours in the case. Why? Because “lessons were learned as a result of this case, and that alone is a benefit to the company.” Besides, the company’s lawyers indicated that the company would not object to paying such a fee, as long as Musk got back his $56 billion in options.

Some observers have suggested that the decision is a clever political compromise: Musk gets his $56 billion, the Court of Chancery is spared a humiliating reversal, and the plaintiff lawyers get a payday that, while tremendous by most people’s standards, is tolerable for a trillion-dollar company like Tesla.

Cynics interpret the same facts differently: afraid of the market, the Delaware Supreme Court gave Musk his options; afraid of the judges on the Court of Chancery, it declined to administer a strong rebuke; and afraid of the plaintiffs’ bar, it made sure the lawyers got paid handsomely.

Whichever interpretation one prefers, the one thing no one could say is that the decision holds up logically. It is the job of the Delaware Supreme Court to administer the state law of fiduciary duties—and that would require finding that the transaction-specific controller doctrine was unsupportable.

Given that mistake, the Supreme Court was right to overrule the Court of Chancery’s decision on rescission. But once it blamed the plaintiffs lawyers for not requesting the correct remedy and awarded only nominal damages of $1, there was no coherent argument for awarding the plaintiff’s lawyers a fee of $54 million. This move seems impossible to defend intellectually.

There may, however, be a bright spot in all this chaos. Since the case was brought as a derivative action, Musk and the directors must pay the $1 in nominal damages not to the shareholder who brought the suit, but to the company, for the benefit of all the shareholders. This means that, technically speaking, the company is the client of the plaintiff’s lawyers. Since the Supreme Court held that this client could get only $1 in damages precisely because the lawyers failed to request the proper remedy (the implication being that the company could have recovered much more if the lawyers had done their jobs right), it would seem that the client could sue the lawyers for malpractice. The client, Tesla, is run by Elon Musk—and he’s just the guy to bring such a suit.

Photo by BRENDAN SMIALOWSKI/AFP via Getty Images

• SpaceX IPO Timing: Musk aims to complete SpaceX’s IPO by July to support ambitious expansion efforts.
• AI Competition Pressure: Competition from OpenAI and Anthropic eyeing IPOs heightened urgency for SpaceX to go public.
• Orbital Data-Center Push: Breakthroughs in computing nodes led SpaceX to prioritize deploying solar-powered AI data centers in orbit.
• Capital Need for Satellites: Thousands of orbital AI satellites will require tens of billions in funding, motivating pursuit of IPO capital.
• xAI Support Strategy: SpaceX’s public valuation seen as potential funding source for Musk’s AI venture xAI and to secure a customer for orbital data services.
• Intercompany Ties: SpaceX invested $2 billion into xAI and already uses xAI technology for Starlink support.
• Internal Communication: CFO Bret Johnsen informed employees that space-based AI data centers were a primary IPO rationale.
• Starship Dependence: Successful orbital data centers rely on Starship, which remains in testing without operational payloads yet.

By

Corrie Driebusch

,

Emily Glazer

and

Micah Maidenberg

4

SpaceX conducted a test flight of its Starship rocket in October. Eric Gay/AP

SpaceX resisted going public for years. Then came the rise of artificial intelligence.

Elon Musk’s rocket maker became one of the country’s most valuable private companies due in part to its ability to develop risky space businesses outside the scrutiny of public investors. Its executives liked to say the company wouldn’t IPO until its rockets were regularly flying to Mars.

That was before the rush to build data centers for AI computing prompted Musk, Jeff Bezos and others to propose putting them in space. The idea has prompted skepticism from many engineers, given the technical challenges posed by building solar-powered AI data centers that zip around Earth.

But it has continued to gain traction and Musk has become obsessed with the idea of SpaceX being the first to do it, people familiar with the matter said. Such a feat would be hard to attempt without the billions of dollars in capital an initial public offering could deliver in one fell swoop.

The billionaire also sees a SpaceX IPO as a way to help his AI company, xAI, catch up to rivals, some of the people said. Musk has a long-running rivalry with OpenAI Chief Executive Sam Altman, who last year explored buying a rocket company to deploy satellites with AI computing capabilities into space.

Two of xAI’s competitors, OpenAI and Anthropic, are eyeing their own IPOs this year, and Musk seems eager for SpaceX to hit the public market first, some of the people said.

SpaceX is expected to select banks to lead the stock offering soon; Musk has told people he wants to complete the IPO by July. 

Elon Musk has told people he wants to complete the IPO by July.  Evelyn Hockstein/Reuters

Musk’s apparent change of tune on SpaceX’s IPO plans in the middle of last year surprised many. The billionaire has loudly complained about running his existing public company, Tesla, and repeatedly tangled with regulators and the courts over issues like his compensation.

SpaceX had for years been developing technology such as computing nodes that would be helpful for an AI satellite network, former employees say. By this fall, it had a breakthrough in its attempts to figure out how to build and launch data centers into space, according to people familiar with the matter, after devoting more resources toward solving technical issues.

But competitors were pushing forward, too. Altman investigated partnering with or acquiring Stoke Space, a startup rocket maker, over the summer. During an event in Italy in October, Bezos said that shifting data centers to orbit made sense. 

Later that month, Musk took to his social-media platform, X, and suggested that solar-powered AI satellites were the future. He posted about the topic several more times in November.

What had been a medium- or long-term goal for SpaceX—developing orbital data centers—became a matter of utmost urgency for Musk mid-last year, according to people familiar with the matter.

Building and launching thousands of satellites would be technically demanding and costly, and SpaceX officials quickly decided the easiest way to raise the tens of billions of dollars it required was to dip into the newly thawed U.S. market for IPOs, the people said.

A data center built by Elon Musk’s xAI in Tennessee. Kevin Wurm for WSJ

Meanwhile, Musk’s AI company, xAI, was trailing rivals including OpenAI and Google’s Gemini by key metrics such as revenue and user base. A SpaceX IPO was seen by some of its investors as a potential cash cow that could supercharge xAI’s growth and in turn help SpaceX.

If SpaceX succeeds with putting data centers in space, its investors expect xAI to become a customer.

Some investors believe SpaceX could purchase a percentage of xAI, or Musk, who owns more than 40% of SpaceX, could tap that stake to invest in his other companies, including xAI. Plus, a publicly traded stock creates a capital safety net of sorts and would be reassuring to investors.

The two companies have existing ties: In July, SpaceX agreed to invest $2 billion into xAI, The Wall Street Journal has reported, as part of a $5 billion financing. The xAI chatbot Grok powers customer-support features for Starlink, SpaceX’s satellite internet service.

In early December, SpaceX Chief Financial Officer Bret Johnsen began privately telling some investors that the company was considering an IPO in 2026; later that month banks began their bake-off pitches to the company to lead the stock offering. 

Johnsen sent SpaceX employees a memo Dec. 12, noting that a key reason the company was investigating an IPO was to “deploy AI data centers in space.”

SHARE YOUR THOUGHTS

Is this summer the right time for SpaceX to go public? Join the conversation below.

Getting satellites for AI launched will be no easy feat. SpaceX will need to get its Starship vehicle working, as its orbital data-center satellites are based on a design optimized for the rocket, the Journal has reported. 

The company has been testing the powerful rocket in flights for almost three years, but it hasn’t deployed any operational payloads yet. SpaceX is expected to launch an upgraded Starship on a test flight soon.

Write to Corrie Driebusch at corrie.driebusch@wsj.com, Emily Glazer at Emily.Glazer@wsj.com and Micah Maidenberg at micah.maidenberg@wsj.com

Copyright ©2026 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8

What to Read Next

[

SpaceX Spat Ends in Defeat for Activist Boaz Weinstein

](https://www.wsj.com/finance/investing/spacex-boaz-weinstein-hedge-fund-investment-a4fb1309?mod=WTRN_pos1)

[

The hedge-fund rabble-rouser failed to wrest control of a U.K. investment trust prized for its valuable SpaceX shares.

](https://www.wsj.com/finance/investing/spacex-boaz-weinstein-hedge-fund-investment-a4fb1309?mod=WTRN_pos1)

Continue To Article

EXCLUSIVE

[

Elliott Invests in Hypersonic-Flight Company Stratolaunch

](https://www.wsj.com/business/deals/elliott-invests-in-hypersonic-flight-company-stratolaunch-db379965?mod=WTRN_pos2)

[

The activist hedge fund is also getting board representation at the Cerberus-backed company.

](https://www.wsj.com/business/deals/elliott-invests-in-hypersonic-flight-company-stratolaunch-db379965?mod=WTRN_pos2)

Continue To Article

[

Tech, Media & Telecom Roundup: Market Talk

](https://www.wsj.com/tech/tech-media-telecom-roundup-market-talk-71bd8506?mod=WTRN_pos4)

[

Find insight on Netflix, Constellation Software and more in the latest Market Talks covering Technology, Media and Telecom.

](https://www.wsj.com/tech/tech-media-telecom-roundup-market-talk-71bd8506?mod=WTRN_pos4)

Continue To Article

[

Tensions Over Greenland Rattle Wall Street

](https://www.wsj.com/finance/stocks/global-stocks-markets-dow-news-01-20-2026-e509b702?mod=WTRN_pos5)

[

Stocks, bonds and the dollar slumped after tensions over Greenland and trade threats. The Nasdaq composite slid 2.4% in its worst day since October.

](https://www.wsj.com/finance/stocks/global-stocks-markets-dow-news-01-20-2026-e509b702?mod=WTRN_pos5)

Continue To Article

[

Why the Tech World Thinks the American Dream Is Dying

](https://www.wsj.com/tech/ai/why-the-tech-world-thinks-the-american-dream-is-dying-daf793dc?mod=WTRN_pos6)

[

Silicon Valley fears this is the last chance to amass generational wealth before AI makes money worthless.

](https://www.wsj.com/tech/ai/why-the-tech-world-thinks-the-american-dream-is-dying-daf793dc?mod=WTRN_pos6)

Continue To Article

[

Supersize CEO Pay Packages Aren’t Paying Off for Shareholders

](https://www.wsj.com/business/ceo-pay-package-issues-f8fc7175?mod=WTRN_pos7)

[

The prospect of a nine-figure paycheck was supposed to spur CEOs to deliver outstanding results, but few of these Elon Musk-like experiments are working out that way.

](https://www.wsj.com/business/ceo-pay-package-issues-f8fc7175?mod=WTRN_pos7)

Continue To Article

[

I got hate mail for being a Tesla stock bull — but here’s the No. 1 thing many are missing about Elon Musk’s AI ambitions

](https://www.marketwatch.com/story/i-got-hate-mail-for-being-a-tesla-stock-bull-but-heres-the-no-1-thing-many-are-missing-about-elon-musks-ai-ambitions-44643c65?mod=WTRN_pos8)

[

Tesla stock is the only portal right now to xAI and Musk’s other empire-building plans.

](https://www.marketwatch.com/story/i-got-hate-mail-for-being-a-tesla-stock-bull-but-heres-the-no-1-thing-many-are-missing-about-elon-musks-ai-ambitions-44643c65?mod=WTRN_pos8)

Continue To Article

[

One of the Largest Units at a Billionaire Haven in Miami Beach Lists for $25 Million

](https://www.mansionglobal.com/articles/one-of-the-largest-units-at-a-billionaire-haven-in-miami-beach-lists-for-25-million-3c05f330?mod=WTRN_pos9)

[

The condo was once the personal residence of Alan Faena, who developed the namesake Faena House

](https://www.mansionglobal.com/articles/one-of-the-largest-units-at-a-billionaire-haven-in-miami-beach-lists-for-25-million-3c05f330?mod=WTRN_pos9)

Continue To Article

Videos