Michael Saylor: How ChatGPT Helped Rake in $15 Billion in a Year?

marsbitPublished on 2026-08-10Last updated on 2026-08-10

Abstract

Michael Saylor, the executive chairman of Strategy (formerly MicroStrategy), claims to have raised $15 billion in the past year using artificial intelligence, specifically ChatGPT. This capital was not personal profit but funds secured through innovative financial instruments to purchase more Bitcoin. With MicroStrategy's traditional financing avenues exhausted, Saylor turned to ChatGPT to design a novel class of securities: variable-dividend preferred stocks. These products, including the STRC, STRK, and STRF, were engineered to provide stable funding without the volatility of common stock or the maturity pressure of debt. Despite significant skepticism from bankers and lawyers, these AI-conceived structures were successfully launched, fueling MicroStrategy's continued massive Bitcoin acquisitions. While the company's common stock value has fallen roughly 80% alongside Bitcoin's price, Saylor defends the strategy as a masterclass in using AI to execute unprecedented financial engineering. The move has drawn both sharp criticism and praise, solidifying Saylor's reputation for leveraging new tools to advance his Bitcoin-centric vision.

Written by: Billy Bambrough

Compiled by: AididiaoJP

The best storyteller in the Bitcoin space has once again left everyone speechless.

Michael Saylor, Executive Chairman of Strategy (formerly MicroStrategy), whose net worth fluctuates wildly with Bitcoin's price. According to Forbes' latest estimate, his current net worth is approximately $4 billion. In 2020, he transformed this originally enterprise software company into a Bitcoin hoarding machine, and his wealth curve has since almost perfectly followed the coin's price.

Over the past six years, Strategy has accumulated nearly 850,000 Bitcoins, with a current market value slightly over $50 billion. The company's Bitcoin reserve once approached $55 billion. These coins were largely not purchased with profits from core business operations but were acquired through continuous issuance of debt, equity, and preferred stock.

Now, Saylor attributes the most crucial fundraising round of the past year—approximately $15 billion—directly to artificial intelligence.

On August 6th, he said to host Steven Bartlett on the popular podcast *The Diary Of A CEO*: "I used AI to make $15 billion last year."

The crypto world instantly exploded.

Saylor immediately clarified: This $15 billion is not personal profit he pocketed, nor is it company book profit. It is the actual capital raised from the market through a series of Bitcoin-backed preferred stock products. This money was almost entirely used to buy more Bitcoin.

He broke down the numbers: One preferred stock product alone raised about $10.5 billion ($2.5 billion in the initial IPO, with subsequent shelf offerings selling another $8 billion), while several related products contributed a combined total of about $4 billion, adding up to roughly $15 billion.

This is the part that truly stunned both Wall Street and the crypto community—these preferred stock products were essentially "conversed" into existence by Saylor and his team, piece by piece, using ChatGPT.

The Traditional Financing Path Reached Its End

By early 2025, Strategy had become one of the world's largest issuers of convertible bonds. The convertible bond route was largely exhausted, and continuing to issue common stock would severely dilute shareholders. The company already held Bitcoin worth about $30 billion, but to continue large-scale accumulation, it needed to find a completely new financing instrument.

Saylor didn't call for another investment bank brainstorming session. He simply opened ChatGPT.

He later recalled: "We went and asked AI: Can you design a security that sits between common stock and traditional debt? AI said, of course you can, you do this, that, and the other thing..."

ChatGPT helped him outline a structure never truly implemented before—a Variable Rate Dividend Preferred Stock. The core design: the dividend rate could be flexibly adjusted monthly (later changed to bi-weekly), with the goal of keeping the preferred stock's market price consistently stable around its $100 par value. This way, it would neither fluctuate wildly like common stock nor have the repayment pressure of traditional bonds, while still continuously funneling capital to the company for buying coins.

Bankers and lawyers initially objected collectively: "No one has ever done this in history."

Saylor's response was hardcore: "Is it illegal? No. Why hasn't anyone done it? Because no one previously had a reason to do it."

AI not only provided the structural design but also helped the team devise responses to regulatory and legal challenges. Ultimately, these products launched successively in the first half of 2025.

The core products were:

  • STRK (Perpetual Strike Preferred Stock): 8% fixed dividend, payable in cash or common stock, convertible to common stock, offering some upside potential.
  • STRC (Perpetual Stretch Preferred Stock): Variable dividend (currently set at an annualized 12%), paid bi-weekly, designed specifically to keep the price as close to the $100 par value as possible, targeting stable income.
  • STRF (Perpetual Strife Preferred Stock): Non-convertible, more fixed-income-like, higher priority, targeting conservative investors.

STRC was the most controversial and also the largest fundraising product. It fell from near par value to around $75 by late June but has rebounded strongly in recent weeks, approaching the $100 target price again. Saylor's team has even publicly stated that "getting STRC back to the $99-$100 range" is one of their most important current operational goals.

Common Stock Halved While Company Hoards Cash

While preferred stock fundraising surged ahead, Strategy's common stock crashed alongside Bitcoin's price. Over the past year, the common stock's market value evaporated by about 80%. The company also began actually selling Bitcoin—the latest public disclosure showed a sale of 1,638 coins in one week, netting about $104.7 million, used partly to pay preferred stock dividends and partly to repurchase its own preferred stock.

Saylor's explanation for this is blunt: on one hand, it's "vaccinating the market," getting people accustomed to the idea that Strategy will also sell coins; on the other hand, it's to genuinely cover debt and dividend payments.

Currently, besides Bitcoin worth approximately $55 billion, the company has amassed around $4 billion in cash reserves. According to management, this sum is sufficient to cover approximately two years of dividend and interest payments, eliminating the need for frequent fundraising.

Some Call it "Terrifying," Others Praise it as "Genius"

After the news broke, tech investor and All In podcast host Jason Calacanis fired off on X: "Is this a terrifying confession from Saylor or a brilliant use of AI?"

But others stand on the opposite side. Bill Barhydt, CEO of the Bitcoin and crypto wealth platform Abra, immediately replied: "History will show this to be an absolutely brilliant piece of financial engineering at the tail end of the US debt cycle. If you can borrow dollars and put them into a fixed-supply asset with fast adoption, you should. Saylor can't legally invest this money into securities (as he's not a company under the Investment Company Act of 1940), so that leaves Bitcoin and gold. He will certainly make mistakes, but the strategy itself is smart."

Saylor himself elevated the matter to a methodological level: "Don't learn the things AI already knows how to do. What you really need to learn is how to get AI to do something no one has ever done before. To achieve earth-shattering success, you have to find that magical opportunity."

This isn't the first time he's publicly thanked ChatGPT. In May 2025, at the company's developer conference, he mentioned using ChatGPT's deep research mode to design the company's convertible preferred stock products.

AI + Bitcoin: Saylor is Writing a New Playbook

From transforming the company into a Bitcoin vault in 2020 to reinventing financing instruments with AI in 2025, Saylor has once again proven: in the crypto world, the most valuable thing is never Bitcoin itself, but the people daring enough to use new tools to amplify it.

$15 billion wasn't money directly spat out by AI. It was a traditional entrepreneur using a new tool to pry open a crack in traditional financial rules, then funneling money ceaselessly into Bitcoin.

As for where this money will ultimately take Strategy and Bitcoin, no one dares to conclude yet. But at least in the summer of 2026, Saylor has once again firmly fixed everyone's attention on "what AI can help people do that has never been done before."

Related Questions

QAccording to the article, how much did Michael Saylor claim to have raised in the past year using AI, and what was the exact nature of this capital?

AMichael Saylor claimed to have raised approximately $15 billion in the past year using AI. This capital was not personal profit or company earnings, but funds genuinely raised from the capital market through a series of Bitcoin-backed preferred stock products. Nearly all of this money was used to continue buying Bitcoin.

QWhat specific financial product did ChatGPT help Michael Saylor design, and what was its key feature?

AChatGPT helped Michael Saylor design a novel financial structure called 'variable dividend preferred stock'. Its key feature was that the dividend rate could be flexibly adjusted monthly (later bi-monthly), with the goal of keeping the market price of this preferred stock consistently around its $100 par value. This design aimed to provide stability, avoiding the high volatility of common stock and the repayment pressure of traditional bonds.

QName the three main preferred stock products mentioned in the article that resulted from the AI-assisted design process.

AThe three main preferred stock products mentioned are: 1) STRK (Perpetual Strike Preferred Stock): 8% fixed dividend, payable in cash or common stock, convertible to common stock. 2) STRC (Perpetual Stretch Preferred Stock): Variable dividend (currently adjusted to 12% annualized), paid bi-monthly, designed for price stability near $100 par. 3) STRF (Perpetual Strife Preferred Stock): Non-convertible, higher priority, more fixed-income oriented for conservative investors.

QWhat contrasting reactions did the news of Saylor's AI-assisted $15 billion raise elicit from notable figures in the tech and crypto space?

AThe news elicited contrasting reactions. Tech investor Jason Calacanis criticized it on X, calling it 'Saylor's horrific admission' and questioning whether it was a brilliant use of AI. Conversely, Bill Barhydt, CEO of Abra, praised it as 'an absolute masterclass in financial engineering' at the tail end of the US debt cycle, arguing the strategy of borrowing dollars to invest in a fixed-supply asset like Bitcoin was smart.

QWhat core advice does Michael Saylor give regarding the use of AI for achieving significant success, based on his experience described in the article?

AMichael Saylor advises that one should not learn to do things AI can already do. Instead, the key is to learn how to make AI do something that has never been done before. He states that to achieve earth-shattering success, one must find that 'magic opportunity' where AI can be applied to create something entirely novel.

Related Reads

Moore Threads, Having Just Raised 8 Billion, Is Already Planning a Hong Kong IPO

On August 9th, Moore Threads announced plans for a Hong Kong IPO alongside a strong first-half 2026 earnings report. Listed on the Shanghai STAR Market in December 2025 as the "first domestic GPU stock," the company raised 8 billion yuan. Its stock price initially surged but has since fallen 40% from its peak. This decline followed market concerns over its use of substantial idle raised funds for low-risk financial products instead of promised projects. The company's H1 2026 revenue reached 1.74 billion yuan, a 147.4% year-over-year increase, already surpassing its full-year 2025 revenue. This growth is attributed to strong demand for AI and full-feature GPUs, particularly its Kuae AI computing clusters. Despite the revenue surge, Moore Threads remains unprofitable on a non-GAAP basis, with a net loss of 115 million yuan after adjustments. R&D and sales expenses grew significantly. The proposed Hong Kong listing aims to support its global strategy, attract international talent, and improve corporate governance, following the path of other Chinese tech firms pursuing dual listings. However, its港股 IPO valuation may face challenges as international investors could benchmark it against peers like Biren Technology, potentially demanding a discount. In summary, while Moore Threads is expanding rapidly and strategically seeking a港股 listing for global growth, it continues to face profitability challenges amid high costs and market skepticism about its capital allocation and valuation.

marsbit4m ago

Moore Threads, Having Just Raised 8 Billion, Is Already Planning a Hong Kong IPO

marsbit4m ago

Hassabis Predicts: AI's 'Move 37' in Mathematics is Only a Matter of Time

Demis Hassabis, the co-founder of DeepMind, predicts that AI achieving a "Move 37" level breakthrough in mathematics—solving a major unsolved problem like a Millennium Prize problem—is only a matter of time. He draws a parallel to the pivotal "Move 37" in the 2016 AlphaGo vs. Lee Sedol match, where AI's unconventional play challenged human understanding. Conversely, Lee Sedol's "Move 78" victory highlighted human resilience and the potential for human-AI collaboration. The article details AI's rapid progress in mathematics. Starting from basic arithmetic errors in 2023, AI systems like AlphaProof and Gemini Deep Think reached gold-medal levels in the International Mathematical Olympiad by 2025. The shift from problem-solving to genuine research began in 2025, with AI (e.g., GPT-5, Claude Fable 5) contributing key ideas to solve or refute long-standing conjectures like the Erdős problem 848 and the Jacobian conjecture. In 2026, AI achieved breakthroughs such as OpenAI's solution to the plane unit-distance problem and DeepMind's Aletheia agent autonomously tackling multiple open problems. Hassabis envisions a "centaur" era of science, where human intuition guides AI in complex, less-verifiable fields like biology and drug discovery. While AI can enhance human creativity and efficiency, as seen in post-AlphaGo围棋, there is a risk of over-reliance diminishing human judgment. The core challenge is for humanity to find its own "Move 78"—leveraging AI while maintaining critical insight and direction.

marsbit6m ago

Hassabis Predicts: AI's 'Move 37' in Mathematics is Only a Matter of Time

marsbit6m ago

Trading

Spot
活动图片