Ross Ulbricht, founder of the defunct Silk Road darknet marketplace, has raised over $1.3 million worth of Bitcoin through the sale of personal and prison-related items.
The auction, held via Scarce City, coincided with his appearance at the Bitcoin 2025 conference—his first public event since being released from prison earlier this year following Donald Trump’s presidential pardon.
Silk Road Founder Makes Public Comeback
The items up for sale included his prison-issued ID cards, clothing, paintings, and handwritten notes. His 2024–2025 prison ID fetched the highest bid at 5.5 BTC, while the full set of three IDs sold for a combined 7.5 BTC, valued at over $780,000 at the time.
Ross Ulbricht’s Prison Memorabilia. Source: Scarce.city
Other memorabilia included a notebook sold for 1.06 BTC, three prison paintings that brought in a total of 2.41 BTC, and clothing such as his prison sneakers and sweatsuit, which sold for 0.54 and 0.51 BTC, respectively.
Ulbricht also parted with personal belongings from before his arrest, including a djembe drum, a backpack, and a sleeping bag. In a statement shared on the Scarce City auction platform, he explained that these items represent a chapter he is now ready to leave behind.
“I’ve left Arizona, the state where I was in prison. It’s time to travel. That means downsizing and turning the page. I’ve decided to auction some personal items from before my arrest and during my time in prison. I don’t need the reminders and I’m sure some of you will love to have them,” Ulbricht said.
The Silk Road founder also highlighted the values he believes should shape the next phase of crypto development, including freedom, decentralization, and unity.
“When it comes to freedom, we’re not there yet. There’s still more freedom to be won,” Ulbricht stated.
Ulbricht’s presence at the event signals a renewed effort to participate in the crypto space, albeit from a different vantage point.
Coinbase, the largest crypto exchange in the US, has successfully evaded a supply chain attack that could have compromised its open-source infrastructure.
On March 23, Yu Jian, founder of blockchain security firm SlowMist, flagged the incident in a post on X, referencing a report from Unit 42, the threat intelligence division of Palo Alto Networks.
The threat actor forked agentkit and onchainkit repositories on GitHub, inserting malicious code intended to exploit the continuous integration pipeline. The suspicious activity was first detected on March 14, 2025.
“The payload was focused on exploiting the public CI/CD flow of one of their open source projects – agentkit, probably with the purpose of leveraging it for further compromises,” Unit 42 reported.
The attacker exploited GitHub’s “write-all” permissions, which allowed the injection of harmful code into the project’s automated workflow. This method could have enabled access to sensitive data and created a path for broader compromises.
A Malicious Commit Targeting Coinbase. Source: Unit42
However, Unit 42 reported that the payload collected sensitive information. It did not contain advanced malicious tools like remote code execution or reverse shell exploits.
Meanwhile, Coinbase responded quickly, collaborating with security experts to isolate the threat and apply necessary mitigations. This rapid action helped the company avoid deeper infiltration and prevented potential damage to its infrastructure.
Despite the failed attempt, the attacker has since shifted focus to a larger campaign now drawing global attention.
In light of this, SlowMist founder advised developers using GitHub Actions—especially those working with tj-actions or reviewdog—to audit their systems and confirm that no secrets have been exposed.
“If your company uses reviewdog or tj-actions, do a thorough self-examination,” Yu Jian stated on X.
This incident highlights the growing importance of securing open-source tools as the crypto ecosystem expands. Data from DeFillama shows that the crypto industry has recorded exploits of more than $1.5 billion this year.
In 2025, AI agents became the newest obsession for crypto market participants. They were integrated into decentralized finance (DeFi), gaming, infrastructure, and even DAO governance, touted as the next evolution of Web3 intelligence.
With this in mind, BeInCrypto contacted OORT CEO Dr. Max Li for his perspective on whether these autonomous, machine-learning-driven software acting on behalf of users could reshape crypto. Li had some interesting insights, but warned that real-world adoption, security, and regulation are the biggest hurdles ahead.
The AI Agent Gold Rush: Disruption or Distraction?
Data from the AI Agents Directory indicates an average monthly increase of 33% in the number of AI agents.
However, despite the growing interest, Web3-based artificial intelligence solutions still account for a minimal fraction (3%) of the overall AI agent ecosystem.
According to Dr. Max Li, founder and CEO of decentralized cloud network OORT, the space is moving faster than its infrastructure can handle, pointing to models like ElizaOS (formerly ai16z).
Yet, in his opinion, the broader playing field is not ready. He says the core infrastructure, from decentralized storage to tokenized agent marketplaces, is still under construction.
The Real Bottleneck? Security, Not Speed
While scalability is often seen as crypto’s weakness, Max Li says security and compliance are bigger threats. This is especially true when tokenizing AI outputs like computing, decision-making, or real-time data.
Dr. Li added that tokenized AI raises difficult questions. Who owns the data that the agents generate? How can decentralized systems comply with global data laws like GDPR? And what happens when AI agents interact with sensitive personal or financial information on-chain?
“These may already be more significant barriers than scalability,” Dr. Li warned.
The OORT executive emphasized that without clear custodianship or compliance frameworks, the risks extend beyond crypto to regulators, investors, and end-users.
Enterprise Adoption Isn’t Coming Anytime Soon
The industry often claims AI agents will bring real-world industries on-chain. However, Dr. Li says it is still a fantasy, particularly in the public blockchain.
He explained that while enterprises like Walmart could benefit from AI for internal operations, there is little incentive to tokenize those agents. Traditional firms want efficiency and control, not decentralized tokens wrapped around their core systems.
“Most enterprises would prefer to keep that data within their own secured servers rather than exposing it on a public, decentralized network,” he said.
While private chains may offer a bridge, Max Li says the idea of tokenized agents powering real-world logistics or finance is, for now, a crypto-native dream.
A Market Fueled by Hype
AI agent tokens have exploded in 2025. Riding the momentum of both AI and crypto, they have attracted massive capital inflows. However, Dr. Li parallels the dot-com bubble, concluding that while innovation is real, the market is overheated.
Based on this, he does not believe the current rally is sustainable: “It’s fair to say there’s a bubble forming here.”
This sentiment echoes Binance founder Changpeng Zhao (CZ), who recently warned that most AI token projects launch too early.
“Too many AI agent developers focus too much on their token and not enough on the agent’s usefulness. I recommend making a really good agent first,” wrote CZ in a post.
Zhao argued that only a tiny fraction of AI agents, say 0.05%, actually need tokens at this stage. Similarly, Hitesh Malviya, an analyst and popular figure on X, recently echoed this sentiment in a post.
“If you look outside the crypto echo chamber, you’ll find that we do have a solid ecosystem of free and better AI agents—and they don’t have tokens, nor might they ever need one. So, what we’re trading in the name of agents is nothing but memes—a value we created out of thin air, like we always do,” Hitesh observed.
Regulatory Turbulence Ahead
Perhaps the most underappreciated risk in the AI agent boom is regulation. The intersection of open AI systems, tokenized data, and borderless blockchains is a minefield for compliance.
Dr. Li warned of contradictions yet to be resolved: How can decentralized AI be transparent and private? Who is liable when agents act autonomously but cause financial losses?
“In the short term, regulatory intervention will likely create additional hurdles for innovation,” he concluded.
This is especially true where there is no global consensus. Until jurisdictions align on KYC (know-your-customer), AML (anti-money laundering) laws, and data governance, institutional adoption will remain cautious, if not frozen.
While the rise of AI agents is real, their integration into tokenized crypto ecosystems is still a high-risk, high-ambiguity frontier. Infrastructure remains fragile. Legal frameworks are missing, and real-world adoption is still speculative at best.
Dr. Max Li’s view is clear: crypto must shift its focus from hype to functionality—from token-first to agent-first design.
Only then will the next leap in AI-powered decentralization become more than just a market cycle.
