Tag: Business – Decrypt

Coinbase is joining with the Linux Foundation to launch the x402 Foundation, an industry group created to oversee the development of a new internet payments standard.

The foundation will steward the x402 protocol, which lets websites request and receive payments as part of normal web traffic, after Coinbase contributed the technology to the Linux Foundation to place it under neutral governance.

“The internet was built on open protocols,” Linux Foundation Executive Director Jim Zemlin said in a statement. “The x402 Foundation will create an open, community-governed home to develop these capabilities in the open, ensuring they evolve with transparency, interoperability, and broad participation across the ecosystem.”

Under Linux Foundation governance, the x402 protocol will remain vendor-neutral, the Linux Foundation said, allowing for “transparent, community-driven growth, ensuring accessibility, and supporting sustainability.”

Erik Reppel, head of engineering for the Coinbase Developer Platform, said Coinbase will remain involved as a founding participant while the foundation manages development of the technology.

“Coinbase is a founding member and the original creator of the protocol,” Reppel told Decrypt. “Both Coinbase and Base are part of the initial set of industry participants supporting the foundation’s migration to an open-source model. Members will help with governance to help guide the future of x402.”

Launched in 2025 by Coinbase’s developer platform, x402 revived the long-unused HTTP 402 “Payment Required” status code to create a native payment layer for the web, allowing websites and APIs to request payment directly during normal HTTP interactions before granting access to content or services.

Developers have begun experimenting with the protocol as AI agents increasingly perform tasks for users online. Projects, including Sam Altman’s World, are integrating x402 into tools that let agents prove they represent real people, and infrastructure efforts like MoonPay’s Open Wallet Standard are adding support for the protocol.

Companies participating in the launch of the x402 foundation include Google, Stripe, Visa, Mastercard, Shopify, Cloudflare, and the Solana Foundation.

“The shift toward agentic commerce requires cloud infrastructure that is as open as the protocols it supports,” Managing Director, Web3 and Digital Assets at Google Cloud, James Tromans, said in a statement. “By joining the x402 Foundation, Google is reinforcing its commitment to interoperable standards that enable secure, AI-driven transactions across platforms.”

“Solana has been one of the earliest adopters of x402, driving nearly 65% of x402 transaction volume this year, and has a growing ecosystem building products with x402 payments,” Head of AI Growth, Solana Foundation, Rishin Sharma said in a statement. “We’re eager to support the x402 Foundation to build the future of agentic payments and onboard more developers, merchants, and agents to pay-per-request models with stablecoins.”

The x402 Foundation will operate under Linux Foundation governance to support community-driven development of the standard. Organizations that contribute to or use the technology can join the foundation and help guide its development. Early priorities include maintaining interoperability across implementations and supporting developers and merchants building services around the standard.

Shan Aggarwal, chief business officer at Coinbase, said the company views the project as an effort to create a more open infrastructure for the future of online payments.

“Agents are going to buy, sell, and transact on our behalf. They will need a payment rail that’s open, interoperable, and doesn’t require a human clicking confirm purchase,” Aggarwal told Decrypt. “That’s x402, and now it’s governed by the community.”

Google’s open AI ambitions got a lot more serious today. The company released Gemma 4, a family of four open-weight models built on the same research as Gemini 3, and licensed under Apache 2.0—a significant departure from the more restrictive terms on previous Gemma versions.

Developers have downloaded past Gemma generations over 400 million times, spawning more than 100,000 community variants. This release is the most ambitious one yet.

For the past year, the open-source AI leaderboard has been largely a Chinese affair. DeepSeek, Minimax, GLM and Qwen have dominated the top spots, leaving American alternatives scrambling for relevance. As Decrypt reported last year, Chinese open models went from barely 1.2% of global open-model usage in late 2024 to roughly 30% by the end of 2025, with Alibaba’s Qwen even overtaking Meta’s Llama as the most-used self-hosted model worldwide.

Meta’s Llama used to be the default choice for developers who wanted a capable, locally runnable model. That reputation has eroded—Llama’s Meta-controlled license raised questions about its true open-source status, and its performance slipped behind the Chinese competition. The Allen Institute’s OLMo family tried to fill the gap but failed to gain meaningful traction. OpenAI released its gpt-oss models in August 2025, which gave the ecosystem a breath of fresh air, but they were never designed to be frontier competitors.

And yesterday, a 30-person U.S. startup called Arcee AI released Trinity, a 400 billion parameter open model that made a compelling case that the American scene wasn’t completely dead. Gemma 4 follows that momentum, this time with the full weight of Google DeepMind behind it, turning it into arguably the best American model in the open-source AI scene.

The model is “built from the same world-class research and technology as Gemini 3,” Google said in its announcement. Gemma 4 ships in four sizes: Effective 2B and 4B for phones and edge devices, a 26B Mixture of Experts model focused on speed, and a 31B Dense model optimized for raw quality.

The 31B Dense currently ranks third among all open models on Arena AI’s text leaderboard. The 26B MoE sits sixth. Google claims both outcompete models 20 times their size—a claim that holds up, at least against the Arena AI numbers, where Chinese models still hold the top two spots.

We tested Gemma 4. It’s capable, with some caveats. The model applies reasoning even to tasks that don’t require it, which can make responses feel over-engineered for simple prompts. Creative writing is decent—serviceable, not inspired—and likely improves with more specific guidance and prompt engineering.

Where it delivered most clearly was code. Asked to generate a game, the output wasn’t particularly flashy or elaborate, but it ran without errors on the first try. Not bad for a 41 billion parameter model. That zero-shot reliability is arguably more valuable than a prettier result that needs debugging.

You can try the (basic, yet functional) game here.

The four variants cover the full hardware spectrum. The E2B and E4B models are built for Android phones, Raspberry Pi, and edge devices, running completely offline with near-zero latency, native audio input, and a 128K context window. The 26B and 31B models target workstations and cloud deployments, extending context to 256K and adding native function-calling and structured JSON output for building autonomous agents. All four models process images and video natively. The larger models’ full-precision weights fit on a single 80GB NVIDIA H100 GPU; quantized versions run on consumer hardware.

The Apache 2.0 license is the other headline. Google’s previous Gemma releases used a custom license that created legal ambiguity for commercial products. Apache 2.0 removes that friction entirely—developers can modify, redistribute, and commercialize without worrying about Google changing the terms later. Hugging Face co-founder Clement Delangue praised it, saying that “Local AI is having its moment,” and it is the future of the AI industry. Google DeepMind CEO Demis Hassabis went further, calling Gemma 4 “the best open models in the world for their respective sizes.”

That’s a strong claim. Proprietary systems from Anthropic, OpenAI, and Google’s own Gemini still lead on the hardest benchmarks. But for open-weight models you can run locally, modify freely, and deploy on your own infrastructure? The competition just got significantly thinner. You can try Gemma 4 now in Google AI Studio (31B and 26B) or Google AI Edge Gallery (E2B and E4B). Model weights are also available on Hugging Face, Kaggle, and Ollama.

AI systems, like ChatGPT or Claude, are known for their intensive energy usage. They need to store data in one place and then process it elsewhere, constantly moving it back and forth. It’s a problem which could now be fixed with new research.

A team of physicists from Loughborough University have designed a device that can process data that is changing over time directly inside the hardware. Traditional systems have relied on software-based methods to do this in the past.

With this new chip, the team of researchers argue that it could be 2,000 times more energy efficient than existing methods.

“This is exciting because it shows we can rethink how AI systems are built,” said Dr. Pavel Borisov, lead author of the study, in a statement. “By using physical processes instead of relying entirely on software, we can dramatically reduce the energy needed for these kinds of tasks.”

Where conventional AI systems are akin to sending documents back and forth between two offices (memory and processor) over and over, with this new chip, it could be like having one smarter office, working on everything in one place.

Brain gain

At the heart of the chip is a memory resistor, a memory chip that remembers past signals. That memory alters how it responds to new signals—in other words, it’s not just following instructions, but learning from history. This is an idea modelled on the human brain.

“Inspired by the way the human brain forms very numerous and seemingly random neuronal connections between all its neurons, we created complex, random, physical connections in an artificial neural network by designing pores in nanometre-thin films of niobium oxide as part of a novel electronic device,” said Dr. Borisov.

“We showed how one can predict the future evolution of a complex time series using these devices at up to two thousand-times lower energy consumption compared to a standard software-based solution.”

AI is often used to process data that changes over time, such as weather reports, stock market tracking or wave analysis. They might not be random, but they are sensitive to small changes.

For these more chaotic kinds of measurements, traditional AI systems need to use huge amounts of energy to keep up with all the small changes, sending information back and forth. This new chip could be perfectly designed for these more chaotic systems.

By analysing past measurements and experiences, the chip better learns to track and understand these chaotic types of measurements, reducing the energy output needed.

While we often think of AI as something like ChatGPT, or facial image software, it is found in most applications these days. This tool is aimed not at static information, like a chatbot, but at time-dependent information.

“Heart beat rates, brain electric activity, the outside temperature. These are all changing every day. There are capable applications that track these, but they are energy-intensive and they require a stable online connection to a server,” Dr. Borisov told Decrypt.

These are the kind of areas this chip could be implemented in, creating smarter systems for data that isn’t stable, often changing throughout time.

“My end goal would be for this kind of technology to be used in a time-dependent signal. Whether that’s in a car, a robot, a nuclear power plant, or in a smart watch,” he added. “For example, to monitor if someone has a stroke or not, to monitor the health of a car engine, or that the nuclear reactor is operating normally, this sort of thing.”

The Justice Department and the CFTC jointly filed lawsuits Thursday against Illinois, Arizona, and Connecticut, coming to the aid of prediction markets in the latest escalation of an ongoing jurisdictional battle over the novel sector.

In the lawsuits, the Trump administration argues the CFTC has exclusive jurisdiction to regulate sports-related wagers on popular prediction market platforms including Polymarket and Kalshi.

In the last year, a growing number of states have sued such platforms, arguing sports-themed prediction markets are not event contracts under the CFTC’s purview—but, instead, unregulated sports betting by another name.

In suing Illinois, Arizona, and Connecticut—three of several states that have come after top prediction market platforms—the Trump administration has made perhaps its most forceful move yet to free the emerging sector from state gambling regulations.

“The CFTC will continue to safeguard its exclusive regulatory authority over these markets and defend market participants against overzealous state regulators,” said CFTC Chairman Michael Selig, in a statement. “This is not the first time states have tried to impose inconsistent and contrary obligations on market participants, but Congress specifically rejected such a fragmented patchwork of state regulations because it resulted in poorer consumer protection and increased risk of fraud and manipulation.”

The DOJ and CFTC argued today that by sending cease and desist letters last year to Polymarket, Kalshi, and Crypto.com, state regulators violated a federal law granting the CFTC exclusive authority to regulate event contracts.

“This Court should put an end to the ongoing efforts by Defendants to undermine the uniform application of federal law,” Thursday’s complaint against Illinois, filed in the U.S. District Court for the Northern District of Illinois, reads.

In the last few weeks, state regulators have notched notable early victories against prediction markets. Last month, Nevada became the first state to successfully ban a prediction market platform temporarily, as its lawsuit against the platform, Kalshi, heads to trial.

Today’s lawsuits could soon be followed by more, given several other states have issued cease and desist orders to prediction market platforms offering sports-related wagers.

In going after Illinois and Connecticut first, the Trump administration has set its sights on two deep-blue states. Trump has, in particular, lambasted Illinois Governor J.B. Pritzker for resisting elements of the White House’s immigration crackdown, and referred to him late last year as “a big fat slob.” Pritzker was named as a defendant in today’s lawsuit against the state of Illinois.

While Arizona did vote for Trump in 2024, the “purple” state currently boasts two Democratic senators and a Democratic governor. Last month, the state became the first in the nation to file criminal charges against a prediction market platform, Kalshi, for allegedly operating an illegal gambling business without a license.

But tensions over prediction market regulation are not strictly partisan. Several red states have gone after Polymarket and Kalshi’s sports wagers in the last few months, including Tennessee. And the government of Utah, which swung for Trump by more than 20 points in 2024, has emerged as one of the leading critics of the ballooning sector.

The Trump administration, however, has taken an aggressively supportive approach towards prediction markets. The president’s media company has its own prediction market-related ambitions, and his son, Donald Trump Jr., currently advises both Polymarket and Kalshi. 

Even the DOJ itself now shares some DNA with the lucrative sector. Yaakov Roth, the principal deputy assistant attorney general representing the federal government in its new case against Illinois, previously represented Kalshi in the company’s landmark victory against the CFTC in 2024.

Ethereum co-founder Vitalik Buterin detailed his personal AI setup in a new blog post, describing the configuration as both “private” and “secure.” Buterin said he runs his artificial intelligence setup entirely on local hardware, and has built custom tools around the large language model (LLM) to prevent his AI agents from sending messages or moving crypto without human sign-off.

“The new two-factor authentication is the human and the LLM,” he wrote.

The post, published Wednesday, marks a step beyond Buterin’s previous calls for privacy-preserving AI. In February, he outlined a four-quadrant Ethereum-AI roadmap spanning private AI use, agent markets, and governance. But this new post goes further, offering a granular look at how he’s actually implemented those principles himself.

Buterin runs the open-source Qwen3.5:35B model locally via llama-server. And after testing multiple setups, he prefers using a  laptop with an Nvidia 5090 GPU that hits 90 tokens per second. That’s fast enough to feel usable, Buterin added.

He stores a full dump of Wikipedia articles and technical documentation on his machine to minimize how often he needs to query external search engines, which he treats as a privacy leak.

The most crypto-relevant disclosure involves how he connects AI to his Ethereum wallet and messaging accounts. Buterin wrote that he built and open-sourced a messaging daemon that allows his AI agent to read Signal messages and emails freely, but restricts outbound messages to himself unless a human manually approves them first.

He advised teams building AI-connected Ethereum wallet tools to adopt the same architecture, with autonomous transactions capped at $100 per day and anything above that requiring confirmation.

The approach is consistent with how Buterin already manages his crypto holdings. He keeps 90% of his funds in a multisig Safe wallet, distributing keys among trusted contacts so that no single person becomes a point of failure.

The AI guardrails appear to be an extension of that same philosophy into an agentic context.

Buterin opened the new blog post by citing security researchers who found that roughly 15% of skills built for OpenClaw, now the fastest-growing GitHub repository in history, contained malicious instructions, with some silently exfiltrating user data without any indication to the user.

“I come from a mindset of being deeply scared that just as we were finally making a step forward in privacy with the mainstreaming of end-to-end encryption and more and more local-first software, we are on the verge of taking 10 steps backward by normalizing feeding your entire life to cloud-based AI,” he wrote in the post.

Coinbase signaled on Thursday that it plans to refine its services after receiving conditional approval for a national trust bank charter from the Office of the Comptroller of the Currency (OCC).

In a blog post, the San Francisco-based crypto exchange said that it is not becoming a commercial bank. Rather, the charter will provide the firm with “federal regulatory uniformity” when it comes to custodying various types of assets on behalf of customers, it said.

The development underscores how crypto-native companies are becoming increasingly tied to the traditional financial system. Coinbase said that the shift will enable the exchange to create new products that cater to both individuals and institutions.

Coinbase highlighted payments as an area that the charter will allow it to expand into. The company already has a deep relationship with stablecoin issuer Circle, which gained approval for a national trust banking charter alongside several competitors last year.

Coinbase Custody Trust Company obtained a limited purpose trust charter from the New York Department of Financial Services in 2018. That established the firm as a qualified custodian, allowing it to safeguard securities on behalf of professionals like investment advisors.

In the blog post, Coinbase said its work with the Department served as a cornerstone toward building operational maturity and institutional trust. The firm said that it will continue to operate under the regulator’s supervision and its stringent BitLicense framework.

Coinbase made clear that it doesn’t plan on accepting deposits from individuals like a traditional bank. What’s more, it doesn’t expect to engage in fractional reserve banking.

Operating under the oversight of a federal regulator, the OCC’s charter removes potential barriers to the exchange’s interstate expansion when it comes to the banking realm. Still, Coinbase’s exchange is already available across the U.S. in all 50 states.

Under the GENIUS Act, a federal framework for stablecoins signed into law last year, the OCC has supervisory authority over qualified stablecoin issuers. The legislation recognizes national trust banks as those permitted to become part of that group. What’s more, the GENIUS Act allows stablecoin issuers to operate without navigating a patchwork of state-level licenses.

When Circle received OCC approval last year, the Office’s Comptroller of the Currency, Jonathan V. Gould, declared that “new entrants into the federal banking sector are good for consumers, the banking industry and the economy.”

Ripple, BitGo, and Paxos Trust Company and Fidelity Digital Assets received a green light at the same time, underscoring how stablecoins are reshaping corporate structures for crypto-native firms. In 2021, Anchorage Digital became the first federally chartered digital asset bank, being awarded that status under the Biden administration.

Not long after the OCC’s wave of approvals last year, Sen. Elizabeth Warren raised the alarm regarding World Liberty Financial’s efforts to attain a charter. She argued in a statement that Gould’s refusal to delay review for the President Donald Trump-banked crypto firm amounted to a “sham.”

“We have never seen financial conflicts of this magnitude and no crypto market structure legislation should pass Congress without guardrails to stop this kind of corruption,” she added.

The Bank Policy Institute, a trade group representing America’s largest banks, urged the OCC to reject a spate of charter applications in October. The organization argued that allowing crypto-native firms to offer bank-like products without greater supervision could “blur the statutory boundary of what it means to be a ‘bank,’” while heightening systematic risk.

Still, the crypto industry’s growing bondage to the traditional financial system has shown no signs of slowing down. In early March, Kraken secured approval for a Federal Reserve “master account,” giving the crypto exchange access to the Fed’s core payment services.

Editor’s note: This story was updated after publication with additional details.

If regulators fail to set up clear policies and rules around prediction markets, there could be another FTX-like implosion, according to Commodity Futures Trading Commission Chairman Michael Selig.

“The failure of agencies to actually regulate, to do their job and set policy, is such a disservice to the builders and innovators and everyday Americans who want to access these products,” Selig said in an interview with Dastan President and Co-Founder Farokh Sarmad.

The CFTC, which claims complete jurisdiction over prediction markets and derivatives contracts, wants to make sure that’s not the case moving forward.

“We saw FTX and we saw all these implosions of crypto firms, I’m concerned we’ll see the same with prediction markets if we keep pushing it offshore into the unregulated space,” he said.

“We’ve got to make sure these exchanges come and register here in the United States and that our rules are set up to facilitate fair markets, markets that have investor protections, customer protections, and have real guardrails and rules,” Selig added.

Prediction markets like Kalshi and Polymarket have exploded in popularity over the last year, moving from niche homes to political markets to massive platforms facilitating event contracts for nearly anything you could imagine—sports, weather, geopolitical issues, and more.

(Disclaimer: Decrypt’s parent company, Dastan, operates Myriad, a prediction market).

The surge in volumes, which now stretches above $20 billion monthly, has led to rapidly rising valuations as well, with Kalshi reportedly doubling its valuation earlier this month when it raised $1 billion at a $22 billion valuation.

Prediction markets under the microscope

Those surging values have not come without scrutiny. Kalshi and Polymarket have recently faced accusations of insider trading on their platforms, as public officials like California Governor Gavin Newsom allege that individuals close to the Trump administration have profited from insider knowledge using prediction markets.

More specific examples have led to a fine and suspension against a video editor for MrBeast, who profited from inside information related to the YouTube personality’s videos. In February, two Israelis were arrested and charged with misusing classified information, having allegedly used military secrets to profit on Polymarket.

Recently, both offered new policies and procedures to try and address the insider trading concerns.

The platforms are also running into issues in the courts, where states are now pushing back on their offerings. For example, Arizona attorney general Kris Mayes filed 20 criminal charges against Kalshi, alleging the site is running an “illegal gambling operation.” 

Earlier this month, Nevada earned a victory in court over the platform, landing a temporary restraining order that banned it from offering its event contracts in the state. Other states, like Massachusetts, may not be far behind from doing the same.

Selig, who has previously condemned the behavior from states, didn’t expect it.

“I am surprised about it,” he told Sarmad about the state’s lawsuits against prediction markets, adding that, “I think the jurisdiction of the agency is very clear.”

The jurisdiction and regulation of prediction markets, as well as crypto and artificial intelligence and their relationship to derivative markets, is something that Selig said the agency is committed to getting right.

States “regulating by litigation”

“We’ve got to get the policy right, and we’re committed to working with every stakeholder who wants to work with us,” he said. “What we don’t want is states and others suing our registrants to assert their authority in the same way that the last administration did with crypto, where they’re regulating by litigation and enforcement actions.”

“We’d like to work together to get the policy right,” he added.

To do so, his agency has recently put out an Advanced Notice of Proposed Rulemaking to invite public comments and feedback about the rules necessary to properly regulate prediction markets.

Last week, the CFTC launched an Innovation Task Force to help create a clear regulatory framework for AI, crypto, and prediction markets as well, indicating it would coordinate with other federal agencies like the SEC and its Crypto Task Force on initiatives.

Global asset manager Franklin Templeton announced Wednesday it is forming a crypto unit—Frankling Crypto—following its acquisition of 250 Digital, a crypto investment firm and spinoff from CoinFund.

The acquisition, expected to close during Q2, brings the 250 Digital team and all its crypto strategies to Franklin Templeton, which will invest in those strategies as part of the deal. 

“Crypto’s institutional moment has arrived, and Franklin Crypto will help our global clients navigate this complex and rapidly evolving asset class by delivering the expertise, knowledge and digital asset products that meet their sophisticated investment needs,” Head of Franklin Crypto Christopher Perkins said in a statement.

Perkins, who will join the firm and serve as the division leader of Franklin Crypto, previously led 250 Digital alongside Seth Ginns, who will come on board and serve as the Chief Investment Officer of Franklin Crypto.

“This is an exciting addition for Franklin Templeton, and we’re pleased to welcome Chris, Seth and the 250 Digital team to our firm,” Franklin Templeton CEO Jenny Johnson in a statement.

“Together, their investment talent and differentiated strategies strengthen our capabilities in digital assets and position us among a small group of global asset managers with a dedicated, institutional-grade crypto investment management team, enhancing our ability to serve clients worldwide,” she added. 

While financial details of the acquisition were not disclosed, the firm noted that BENJI tokens—those representing its on-chain money market fund—will be used as payment considerations, “marking an important and innovative step toward conducting M&A transactions on chain.”

The firm first debuted the tokenized fund in 2021, before expanding its access to Ethereum in 2024. Since that time, it has also expanded to other popular blockchain networks like Solana, and Ethereum scaling networks like Base and Arbitrum. 

Its latest crypto initiative comes just one week after it announced a deal with Ondo Finance to launch tokenized versions of five of its exchange-traded funds (ETFs). 

A representative for Franklin Templeton did not immediately respond to Decrypt’s request for comment.

A security researcher discovered a critical vulnerability in Zcash nodes that could have allowed malicious miners to drain more than 25,000 ZEC from the network’s deprecated Sprout shielded pool—a sum worth about $6.5 million at writing.

Alex “Scalar” Sol disclosed the flaw on March 23, according to a disclosure report released Tuesday, revealing that zcashd nodes were skipping proof verification for transactions involving the legacy Sprout pool. The bug was not exploited and all users’ funds remain safe, according to the disclosure.

The vulnerability spanned releases from July 2020 through the present, with Zcash developers releasing v6.12.0 on Tuesday to contain the fix. Major mining pools moved quickly to patch their systems—Luxor mining pool confirmed deployment on March 25, while F2Pool, ViaBTC, and AntPool all deployed the fix by March 26, according to the same report.

The Zebra full node implementation was not affected by the vulnerability, the report said, and would have triggered a chain fork if exploitation had been attempted, providing an additional layer of network protection.

Sol, who discovered the vulnerability using AI assistance, reported it to Shielded Labs on March 23. The organization coordinated with the Zcash Open Development Lab (ZODL), whose engineer Jack “str4d” Grigg authored the patch.

For his disclosure, Sol will receive a 200 ZEC total bounty—valued above $51,000—with Shielded Labs, ZODL, the Zcash Foundation, and Bootstrap each contributing 50 ZEC.

The Sprout pool was closed to new deposits in November 2020, making it a deprecated but still-active component holding approximately 25,424 ZEC that users have not yet migrated to newer shielded pool versions.

While the vulnerability could have allowed draining these funds, the Zcash Open Development Team (ZODL) said that Zcash’s “turnstile” mechanism would have prevented broader supply inflation. The turnstile requires that any coins leaving the Sprout pool must have verifiably entered it, creating a safeguard against the creation of new tokens beyond the network’s total circulation of around 16.63 million ZEC.

This isn’t the first big vulnerability that the network has faced. Back in 2019, the network patched a bug described as an “infinite counterfeit” crypto generator, though it was patched out before becoming a major issue for the privacy coin network.

Zcash is the biggest gainer over the last 24 hours among the top 100 coins by market cap, per CoinGecko data, rising more than 14% to a recent price above $255. The price of the privacy coin skyrocketed last fall from a price of about $50 to a multi-year peak near $700, but has fallen alongside Bitcoin and other cryptocurrencies in recent months.

Latin America’s e-commerce giant quietly killed its ERC-20 loyalty token four years in—while Nubank’s Nucoin already died the same death.

Mercado Libre is shutting down Mercado Coin. The company—which is bigger than Amazon in Latin America—announced through its Mercado Pago wallet that the token will stop functioning on April 17.

Users holding the coin have three options: Sell it through the app, spend it as purchase credit on the platform, or do nothing and wait for it to be auto-converted to local fiat. Mercado Libre didn’t explain the decision in its customer notice. The company did not respond to Decrypt‘s request for comment.

Mercado Coin launched in August 2022 in Brazil as an ERC-20 token on Ethereum, built along with crypto exchange Ripio. The idea was simple: buy selected products on the marketplace, earn tokens, use them on future purchases, or cash out. It never really caught on.

The token expanded to other markets after its Brazilian debut, but it never built a meaningful base. It lived and died as a curiosity—a loyalty points system with extra steps and a dose of volatility.

That last part is likely why Mercado Libre isn’t abandoning crypto altogether—it’s just changing what kind of crypto it bets on.

In August 2024, the company launched MeliDolar (MUSD), a stablecoin pegged to the U.S. dollar. It’s also built with Ripio, and also accessible through Mercado Pago. The key difference is, of course, the price of MeliDolar doesn’t move. It’s backed by U.S. Treasury securities and dollar deposits, and its value sits exactly where users expect it to.

MeliDolar is also the backbone of the company’s Meli Plus loyalty program. Members get it as cashback on purchases. They can spend it back on the platform, sell it in the app with no fees, or just hold it as a dollar hedge—a genuinely useful feature for Brazilian and Mexican consumers watching their local currencies lose ground. This was the logical step after a brief experiment allowing users to experiment with Paxos’ Stablecoin.

Mercado Libre isn’t alone in learning this lesson the hard way. Nubank—Brazil’s largest bank, a fully digital neobank with over 100 million customers in Brazil alone—ran the same experiment at the same time. Nucoin launched in 2023 on Polygon, airdropped to its massive user base as a loyalty and rewards token.

It subsequently collapsed 97% in value. Nubank suspended trading in September 2024, gave holders 90 days to convert their coins to Bitcoin or USDC, and shut the whole thing down by December. Sixteen million users were left holding the bag, sweetened only by a prize campaign the bank ran to take the edge off.

Mercado Libre still holds more than $38 million in Bitcoin on its balance sheet and continues to offer crypto trading and stablecoin transfers through Mercado Pago. The infrastructure is staying—but not the coin nobody was really using anyway.