Happy Saturday. I'm bringing you the latest from the worlds of Bitcoin, AI, and regulation. Today, Square flips the switch on Bitcoin payments for millions of U.S. merchants by default. Google fires a shot at Meta with Gemma 4, its most capable open-weight AI model family yet. Capital B tokenizes its shares on Liquid, pushing Bitcoin-native financial rails forward. And the SEC greenlights multi-crypto trust options while Morgan Stanley inches closer to launching its own Bitcoin ETF. Let's get into it.
Jack Dorsey's Square just made one of the most quietly significant moves in Bitcoin adoption this year. Instead of asking merchants to opt in to accepting Bitcoin, Square has flipped the model. Bitcoin payments are now automatically enabled for all eligible U.S. sellers, roughly 4 million merchants. If you don't want it, you opt out. That's a massive psychological and practical shift.
Here's how it works. When a customer pays with Bitcoin, the merchant receives U.S. dollars instantly. No volatility exposure, no extra setup, no learning curve. And through the end of 2026, there are zero processing fees. Near-instant settlement. It just works inside the existing Square infrastructure.
Now, why does this matter? Because the biggest barrier to Bitcoin as a payment method has never been technology. It's been friction. Merchants didn't know how to set it up, didn't want to deal with price swings, and frankly didn't see the demand. Square is removing all 3 of those objections in one move. You don't set anything up. You never touch Bitcoin. And you lose nothing by having it on.
The only exclusion right now is New York-based businesses, which isn't surprising given the state's regulatory environment. But for the rest of the country, this is Bitcoin quietly becoming part of the checkout experience at your local coffee shop, barbershop, or farmer's market.
It's worth noting that this fits into a broader pattern at Block. Cash App already lets users buy and send Bitcoin. Block has invested in Bitcoin mining hardware. Dorsey himself has been one of Bitcoin's most vocal advocates. This isn't a marketing stunt. It's a long-term thesis being executed methodically.
The real question is whether consumer demand follows. Merchants are ready now. The infrastructure is there. If even a small percentage of Square's user base starts paying in Bitcoin regularly, the network effects could be significant. This is the kind of move that doesn't make headlines for a week but reshapes behavior over years.
Google DeepMind dropped Gemma 4 this week, and it's a big deal for the open-source AI landscape. This is a full family of open-weight models released under the Apache 2.0 license, which means no commercial restrictions, no usage caps, no strings attached. That's a direct challenge to Meta's Llama, which has dominated the open-weight conversation for the past year.
Gemma 4 comes in 4 sizes, and the range is impressive. The smallest, an effective 2 billion parameter model, requires less than 1.5 gigabytes of memory. You can run it on a Raspberry Pi. The largest supports context windows up to 256,000 tokens and handles images, video, and audio natively. We're talking real-time speech understanding on a phone, without an internet connection.
What makes this strategically interesting is the focus on on-device deployment. Google is betting that the future of AI isn't just cloud APIs. It's models running locally on your phone, your laptop, your smart home hub. This enables use cases where latency matters, where privacy matters, or where connectivity simply isn't available. Think embedded robotics, offline voice assistants, edge computing in industrial settings.
The models also support native function calling and structured JSON outputs, which is critical for agentic workflows. You can build AI that doesn't just generate text but actually takes actions, calls tools, and reasons through multi-step problems. That's the direction the entire industry is headed, and Google is making sure developers can do it with open models.
On benchmarks, Google claims Gemma 4 outperforms larger models in several categories. Take that with the usual grain of salt. Benchmarks are gamed by everyone. But the fact that a sub-2-gigabyte model can compete with much larger architectures is genuinely noteworthy.
The real story here is competitive. Meta has had a near-monopoly on mindshare among developers building with open-weight models. Google releasing under Apache 2.0 with no restrictions is a direct attempt to peel developers away. And for the broader ecosystem, more competition in open AI is unambiguously good. It means better models, fewer lock-in risks, and faster innovation at the edges of what's possible.
Let's talk about something happening on Bitcoin's own financial rails. Capital B, a European Bitcoin Treasury Company listed on Euronext Growth Paris, has completed the tokenization of its existing shares on Liquid, Bitcoin's federated sidechain. This was done through STOKR, a digital asset platform based in Luxembourg that has processed over a billion dollars in tokenized assets.
Here's the structure. Each token, called CALCPB, is backed by 100 ALCPB shares. These tokens are tradable 24/7 in Bitcoin on the Sideswap secondary market, peer-to-peer. No exchange intermediary, no trading hours, no settlement delays. You're trading equity exposure around the clock, settled in Bitcoin, on a Bitcoin sidechain.
This is exactly the kind of use case that Liquid was designed for. Not speculative token launches, but regulated financial instruments issued on Bitcoin-native infrastructure. The tokenization was facilitated through ORO (II) - EQFOUR, a Luxembourg-based securitization fund, so this isn't some Wild West DeFi experiment. It's structured, compliant, and backed by real shares.
Now zoom out. This sits alongside other developments pushing Bitcoin into the role of financial infrastructure. Circle just launched cirBTC, a wrapped Bitcoin token backed 1-to-1 with on-chain reserves, designed to bring Bitcoin into DeFi with more transparency than existing wrapped tokens. And more broadly, we're seeing a trend where Bitcoin-native platforms are starting to compete with Ethereum-based tokenization for serious financial use cases.
The significance of Capital B's move is less about the size of the deal and more about the proof of concept. A listed European company tokenized its equity on Liquid, made it tradable in Bitcoin, and did it within existing regulatory frameworks. That's a template other companies can follow. And as more Bitcoin treasury companies emerge, the demand for this kind of infrastructure will only grow.
Meanwhile, we're also seeing cracks in the corporate Bitcoin treasury model. Genius Group, which targeted a 10,000 BTC treasury in mid-2025, just sold its last 84 BTC to cover $8.5 million in debt. The treasury is now empty. It's a reminder that conviction is easy in a bull market. Execution under financial pressure is what separates strategy from sloganeering.
The SEC approved a rule change this week allowing the NYSE American to list options on multi-cryptocurrency commodity trusts. Until now, options were only permitted on single-asset trusts, meaning you could trade options on a Bitcoin ETF or an Ethereum ETF, but not on a basket product holding multiple digital assets. That just changed.
The requirements are strict. Each cryptocurrency in the trust must have an average daily market value of at least $700 million over the past year. Derivatives on each underlying asset must trade on markets with surveillance-sharing agreements. And the trust shares themselves must meet the same listing standards as traditional ETF options. So this isn't the SEC loosening the reins. It's the SEC extending existing frameworks to cover more complex products.
What this means in practice is that investors will soon be able to hedge or speculate on diversified crypto exposure using options, within a regulated exchange environment. That's a meaningful step for institutional adoption because options are how sophisticated market participants manage risk. Without them, you're limited to directional bets. With them, you can construct nuanced positions.
Separately, Morgan Stanley is inching toward the launch of its own spot Bitcoin ETF. Amendment number 4 was filed with the SEC, and based on the level of detail around custody, operations, and risk disclosures, this looks close to approval. The product, called MSBT, will be listed on NYSE Arca as a passive vehicle tracking Bitcoin's price. No active trading, no leverage, no derivatives.
This matters because Morgan Stanley brings a distribution network that dwarfs most existing Bitcoin ETF issuers. We're talking about a firm with trillions in client assets and a massive wealth management division. BlackRock's IBIT is already processing 16 to 18 billion dollars in daily trading volume, rivaling Binance. Adding Morgan Stanley to the mix intensifies competition and, more importantly, normalizes Bitcoin as a standard portfolio allocation.
But here's the nuance. Despite strong trading volumes, Bitcoin ETFs actually saw net outflows in Q1 2026. Bitcoin had its worst first quarter since 2018. March reversed the trend with net inflows, and April started with modest gains, but the picture is mixed. High volume doesn't always mean new money coming in. A lot of it is hedging and rebalancing. The structural demand is building, but the market is still working through a difficult period.
Here's the thing to sit with this weekend. The infrastructure for Bitcoin to function as both a payment network and a financial platform is being built in plain sight. Square turns on payments for 4 million merchants. Capital B tokenizes equity on Liquid. Morgan Stanley is about to launch an ETF. The SEC is expanding options markets. None of these alone is transformative. But together, they represent a gradual, irreversible integration of Bitcoin into the machinery of everyday finance. The question isn't whether it's happening. It's whether you're positioned for when it becomes unremarkable. Have a good weekend.