Bitcoin’s mining difficulty has fallen below its year-earlier level for only the second time in the network’s history as weak mining economics and the shift toward artificial intelligence weigh on capacity growth.

The metric, which measures how difficult it is to mine a Bitcoin block, is now at 126.23 trillion after falling 0.74%, about 1.1% below the 127.62 trillion reached a year earlier and 19.1% from the 155.97 trillion all-time high seen in November 2025.

Difficulty adjusts every 2,016 blocks, or roughly every two weeks, to keep Bitcoin’s average block time near 10 minutes. Falling difficulty indicates that less computing power was competing during the previous adjustment period, while reducing competition for miners that remain online.

The metric has dropped about 14% from its January peak, reached this year, following declines of 10% in June and 5% earlier in July, according to network data.

The only previous year-over-year decline was after China’s 2021 mining ban, which temporarily removed roughly half of the network’s computing power. Difficulty recovered as miners relocated to other regions.

This time around, the plunge is more mining economics-based.

For years, stablecoins have been marketed as crypto’s breakthrough application for cross-border payments, promising near-instant transfers at a fraction of the cost charged by traditional remittance providers.

Sending USDC across a blockchain may indeed cost only a few cents but a new study from the Bank of Italy suggests that isn’t what most people actually pay when they send money home.

In a mystery-shopping exercise spanning 10 international remittance corridors, researchers found that stablecoin-based transfers were not systematically cheaper than conventional money transfer operators once the full journey, from bank account to crypto wallet and back into local currency, was taken into account.

The study, published as Markets, Infrastructures and Payment Systems Paper No. 86, tracked transfers of 200 USDC from Italy to destinations including Argentina, Brazil, South Africa, the UAE and Japan.

End-to-end costs varied dramatically, ranging from roughly 0.3% to almost 9% of the value transferred depending on the corridor and service providers used. Settlement times also differed widely, from around 20 minutes where domestic instant payment systems supported withdrawals to as long as two business days when recipients relied on conventional bank transfers.

Blind spots

A central bank highlighting shortcoming in the promises that stablecoins may make is in some ways to be expected. Traditional financial (TradFi) institutions may have a vested interest in undermining adoption of stablecoins – digital tokens pegged to fiat currencies. Digital currencies and blockchain were designed to remove much of the need for intermediaries, such as central banks, after all.

An affiliate of a political action committee (PAC) funded largely by contributions from cryptocurrency companies Ripple Labs and Coinbase has poured more cash into ads for next week’s primary race in Michigan’s 13th Congressional District.

According to Federal Election Commission (FEC) filings as of Thursday, the Protect Progress PAC had spent more than a combined $2 million on media to support Michigan Representative Shri Thanedar in the state’s 13th district and oppose his Democratic challenger, Donavan McKinney. 

The most recent filings effectively doubled what the PAC had reported spending a week prior, with an additional $884,240 on ads to support Thanedar and more than $150,000 to oppose McKinney.

Source: FEC

During his time in the US House of Representatives, Thanedar voted in favor of the stablecoin-focused GENIUS Act legislation and the crypto market structure bill currently under consideration in the Senate, the Digital Asset Market Clarity (CLARITY) Act. He also cosponsored the Promoting Innovation in Blockchain Development Act in an effort to protect developers. 

In a July 21 statement on the PAC spending supporting Thanedar, McKinney said “the crypto lobby is paying my opponent back for helping Trump make over $1 billion since taking office.” He was likely referring to the US President disclosing that he earned more than $1.4 billion from crypto investments in 2025, including from his memecoin, Official Trump (TRUMP) and through his family’s business, World Liberty Financial. Many Democrats have accused Trump of using his position to profit from the presidency through laws like GENIUS.

Cointelegraph reached out to Thanedar’s and McKinney’s campaigns for comment on the PAC expenditures but did not receive an immediate response.

Related: US senators sent revised ethics rules to White House for CLARITY Act: Report

Protect Progress is an affiliate of the Fairshake PAC, which was responsible for spending more than $170 million in the 2024 US election cycle through media supporting candidates it considered favoring crypto industry-aligned policies. The Michigan primary is scheduled for Tuesday, but the PAC and its affiliates have already poured millions of dollars into 2026 races in Texas, Illinois and other states.

The US consumer advocacy group Public Citizen reported in June that Fairshake and its affiliates were responsible for spending more than $82 million out of the roughly $189 million crypto companies had used in the 2026 election cycle. Fairshake reported holding a $193 million war chest as of January.

PAC spending in Washington and Alabama with primaries looming

In addition to Michigan’s primaries, the Fairshake affiliate Defend American Jobs PAC spent more than $65,000 on media to support a Republican running in Washington’s 4th congressional, according to FEC filings. Washington is scheduled to hold primaries the same day as Michigan.

Alabama, scheduled to hold primaries on Aug. 11, has also been a focus for Fairshake. FEC filings showed that Defend American Jobs spent more than $511,000 on media to support Jerry Carl Jr., a Republican who represented the state’s 1st congressional district from 2021 to 2025. Notably, the former Alabama lawmaker was one of the wealthiest in the state’s House delegation, with a reported net worth of up to $15 million in 2023.

Magazine: Crypto lobby spending on Republicans far outpaces Democratic support

While exploits in crypto often grab headlines because of the sheer amount of money that gets stolen, Coates emphasized that many of these hacks actually originate outside of blockchain compromises themselves. “In many cases, it is an operational security issue or a Web2 issue that led to a key compromise,” he said.

This will only prove to be more difficult as artificial intelligence advances gives attackers better tools to exploit security practices.

“The social engineering piece is going to get a lot worse because of the power of AI and deepfakes,” Coates said. “We should expect full spoofed phone calls with voices of people that we know… there’s really no reason this won’t hyperscale.

To prevent that, Coates thinks crypto needs to come up with better systems that remain secure and work when people fall for these scams.

“You cannot fully prevent anyone from falling victim,” he said. “Eventually, you will be fooled because the cons are that good.” Organizations should thus have multiple layers of various degrees of security controls, so “when someone gets fooled, the other things take over to protect you.”

For the longer-term, the question of quantum computing largely looms on various crypto ecosystem’s futures, including that of Solana.

The obvious objection is that this is leveraged speculation in new language, and that traditional markets keep their frictions for good reasons. Both points have force. A funding rate is not a substitute for the price discovery that settlement enforces, and continuous leverage on volatile assets concentrates risk in ways periodic markets do not. But those are arguments for building the structure carefully, not for assuming it will not be built. The demand is already here, and it moves toward the venue that offers universal access to global assets, whether stocks, crypto or FX.

There is one place crypto’s progress this cycle has not yet reached. Over the past year, tokens acquired real economic rights, revenue shares, buybacks and votes, while projects with nothing behind them were delisted and some of the strongest teams chose IPOs over token launches. Even the IPO no longer sits outside this system: SpaceX’s shares changed hands as synthetic pre-IPO perpetuals on Hyperliquid for weeks before its June 2026 listing, trading tens of millions of dollars a day in May and swelling to roughly $1.3 billion on debut day as investors shut out of the traditional allocation turned to crypto rails. The market structure beneath perps has not fully followed. The next stage is to build that alignment in, with the transparency now expected across the rest of crypto. That consolidation is happening on centralized venues as much as onchain: the largest exchanges now run multi-asset books where equities, crypto and FX clear side by side, and a single centralized platform accounted for more than half of all real-world-asset perp volume in May 2026. The pull is structural rather than speculative, with 52% of Bitget’s users already holding both stocks and crypto.

Galaxy Research, the research arm of Galaxy Digital, has expanded the on-chain footprint linked to the Coldcard wallet incident that reportedly resulted in the loss of 1,082.65 bitcoin. In a new analysis, the firm identified 1,196 addresses that were involved in transactions tied to that event, widening the estimated scale beyond earlier preliminary figures.

The activity Galaxy Research points to took place between 1:10 AM and 1:51 AM UTC on July 30, spanning blocks 960,183 to 960,191—roughly 30 hours before Coldcard published its first security advisory, as referenced in Galaxy Research’s post on X.

Key takeaways

• Galaxy Research traced a cluster of 1,196 addresses tied to the Coldcard incident, connected to losses of 1,082.65 BTC.
• The movements were observed across blocks 960,183–960,191 between 1:10 AM and 1:51 AM UTC on July 30.
• Galaxy Research says the transactions share a recognizable on-chain pattern (including identical fees and no change outputs), though later attacks may not reuse the same fingerprint.
• Earlier estimates from Rob Hamilton and related analysis suggested a smaller, tighter window of activity that has since been superseded by the larger Galaxy Research mapping.
• Coinkite has said it released a hotfix for a firmware bug, but users who created seeds using the vulnerable firmware may still need to move funds to a new seed.

Galaxy Research widens the address set

According to Galaxy Research, the incident’s impact is visible on-chain in a larger set of wallets than first documented publicly. The firm said it identified 1,196 addresses linked to the Coldcard wallet incident that ultimately resulted in 1,082.65 BTC being moved in the span it analyzed.

Galaxy Research’s tracing work focused on how those funds moved through the network during a specific period. It reported that the key transaction activity occurred between 1:10 AM and 1:51 AM UTC on July 30, crossing blocks 960,183 to 960,191.

That timing is also notable in terms of disclosure. Galaxy Research indicated this took place about 30 hours before Coldcard’s first security advisory was published.

From a smaller estimate to a larger on-chain pattern

Prior to Galaxy Research’s broader mapping, a preliminary view of the incident suggested a more limited sweep. Earlier analysis by AnchorWatch CEO and co-founder Rob Hamilton estimated that 594.48 BTC—worth around $38 million at the time—moved across roughly 500 transactions within a narrow three-block window.

Galaxy Research’s later work does not necessarily contradict the existence of that tight burst; instead, it expands the scope of what can be linked to the event by pointing to a repeatable transaction fingerprint.

In posts on X, Galaxy Research said the identified transactions share specific characteristics, including identical 30 satoshis per virtual byte fees and the absence of change outputs. Galaxy Research described these features as part of the method that allows the initial attack activity to be identified on-chain.

Importantly for users trying to assess exposure, Galaxy Research also cautioned that future attacks targeting Coldcard-generated addresses may not always follow the same on-chain “fingerprint.” That means wallet owners should not assume that the first set of identifiable traits will be reused in later attempts.

What Coinkite says about the firmware bug and the limits of updates

Coinkite co-founder Rodolfo Novak publicly addressed the issue via an X post on Friday, saying the company takes responsibility for the firmware bug and is working to determine the full scope of the problem.

Novak also said Coinkite released a hotfix intended to remove the software fallback path. However, he warned that installing the fix does not retroactively protect seeds that were generated using vulnerable firmware.

In practical terms, Novak advised users who created seeds on the vulnerable firmware to move their funds to a new seed. That distinction—between fixing a flaw going forward and securing already-generated keys—appears to be central to how users should interpret the incident response.

This is also a reminder that “device firmware updates” and “seed security” are not always interchangeable. If the vulnerability affected how seeds were generated or handled, a patch may stop new risk but cannot undo the exposure that may have occurred when the vulnerable firmware produced the original seed material.

Why the expanded tracing matters for incident assessment

The difference between Hamilton’s earlier estimate of 594.48 BTC and Galaxy Research’s later identification of 1,082.65 BTC underscores how incident accounting can evolve as analysts refine clustering techniques and expand time windows. Early on-chain forensics often focus on the clearest bursts; later work may connect additional wallets and transactions using shared traits like consistent fee patterns and transaction structure.

For traders and users, this matters because it changes how incident exposure can be understood. Wallet owners who are evaluating whether they need to move funds may face a moving target: a larger set of addresses suggests that more wallets could have been impacted than initially thought, while Galaxy Research’s warning about fingerprint variability implies that on-chain searches may not capture everything using a single pattern.

For developers and auditors, the episode also highlights the importance of both preventive controls and disclosure timing. Galaxy Research’s observation about the 30-hour gap between the analyzed activity and Coldcard’s first advisory publication frames the timeline in which users may have been acting on incomplete information.

As more details are verified, the key question for the broader ecosystem will be whether subsequent investigations confirm additional waves of activity beyond the identifiable on-chain pattern described by Galaxy Research—and whether Coinkite’s technical findings fully explain how the firmware behavior led to the reported losses.

Readers should watch for further updates from Coinkite on what the bug impacted at the seed level and for additional on-chain analysis that tests whether other clusters of transactions match or diverge from the fee and “no change output” fingerprint outlined by Galaxy Research.