Bitcoin Brief

Sup, freaks.

BIP-110 reached the start of mandatory signaling over the weekend. The code did what its proponents wanted. The overwhelming majority of Bitcoin did not follow it.

Let's get into it.

LEAD STORY

BIP-110 Hit Its Mandatory-Signaling Height. Bitcoin's Hashrate Did Not Follow.

I think it is safe to say the BIP-110 saga has ended with a whimper. Its proponents reached the start of mandatory signaling and watched as miners, economic nodes, exchanges, wallets, and the rest of the market refused to follow their activation path.

At block 961,632, Bitcoin Knots nodes configured to enforce BIP-110 began rejecting blocks that did not signal readiness for the proposal. One important clarification: the underlying content restrictions were not scheduled to become active at that height. Block 961,632 began the mandatory-signaling window. The full rules were scheduled to activate later, at block 965,664, on any BIP-110 chain that advanced that far.

Those rules attempt to restrict several methods of embedding arbitrary data in Bitcoin transactions. Supporters believe the restrictions protect node operators from carrying abusive or illegal data, refocus development on Bitcoin as money, and reduce the ability to use the network for non-monetary purposes.

The proposal violates Bitcoin's content-neutral design. It tries to push subjective policy preferences into consensus and creates a precedent that can lead down a very slippery slope. Once a faction proves it can decide which valid uses are legitimate, every future faction gets to make its own list.

The way BIP-110 proponents tried to achieve consensus was absolutely piss poor from my perspective. The deployment lowered the ordinary BIP9 signaling threshold from 95% to 55%, imposed a mandatory-signaling period, and made lock-in inevitable on any BIP-110-enforcing chain that advanced through that period. Because non-signaling blocks were rejected, a surviving enforcing chain would enter LOCKED_IN at block 963,648 even if voluntary signaling never reached 55%. When the overwhelming majority of participants refused to follow, Luke Dashjr framed the ordinary Bitcoin chain as an attack.

The implementation did receive public discussion and review, but nowhere near the broad, independent scrutiny that a contentious consensus change deserves in my opinion. The BIP's own rationale says the rules were deliberately simplified to minimize review time because supporters considered the change urgent. Urgency is not an excuse to rush consensus code, particularly when the proposal alters Bitcoin's rules for subjective reasons.

Luke Dashjr declared BIP-110 active and framed blocks that violated its mandatory-signaling rules as attacks on Bitcoin. Bitcoin Knots issued its own mandatory-signaling notice. The overwhelming majority of miners kept extending the existing Bitcoin chain. Exchanges, wallets, markets, and the rest of the economic network continued treating that chain as bitcoin. The BIP-110-enforcing branch pulled only negligible hashrate and showed no observable meaningful exchange or market activity.

Individual node enforcement is not fake or useless. Your node decides which blocks and transactions you accept. Nobody can force your machine to recognize a rule set you reject. Bitcoin sovereignty starts there.

But sovereignty is not the same as command.

Your node can enforce any rules you want. It cannot force miners to build on those rules, exchanges to list the resulting asset, wallets to recognize it, or other people to value it. Consensus emerges when independent actors converge on compatible rules and accept the economic consequences of enforcing them. A software release can define a chain. It cannot manufacture a market around that chain.

BIP-110 exposed the difference between cheap signaling and costly enforcement. Supporting a rule in a poll, a node count, or a social post is easy. Remaining on a low-hash branch while the liquid market continues somewhere else is expensive. The mandatory-signaling height forced supporters to reveal whether they were willing to pay that price.

The answer was clear.

Luke Dashjr is now openly discussing a change to Bitcoin's proof-of-work algorithm so the BIP-110 branch can continue without competing against Bitcoin's SHA-256 hashrate. No replacement algorithm has been publicly settled. If that camp moves forward with a new proof-of-work algorithm, it will be a backward-incompatible hard fork with different consensus rules and a need to attract its own miners and market support. At that point there will be no ambiguity. They will have launched an altcoin.

This has been an abject failure. BIP-110-enforcing nodes began mandatory signaling. Bitcoin's dominant chain did not follow.

SIGNAL

BTCPay's Exploit Was an LND Credential Theft

Friday's emergency warning told every BTCPay operator to update. Weekend guidance narrowed the known attack path. BTCPay maintainers say affected LND-backed setups exposed .macaroon credentials that attackers could use to control Lightning nodes and steal funds. The project's current analysis says BTCPay-generated on-chain wallets were not the drain path, but the full postmortem is still pending. If you run BTCPay, install version 2.4.2. If you use LND, regenerate macaroons and rotate other Lightning authentication material. Confirmed losses are real, but the team has not published an authoritative victim count or total. Updating the application closes the known software path. Rotating exposed credentials removes the keys an attacker may already possess.

Bitcoin's Red Team Reached 501 Projects in 108 Hours

Calle's 108-hour situation report says the Bitcoin Red Team expanded to 501 projects, logged 7,958 findings, and classified 1,280 as high or critical after tightening its severity standards. The campaign says 24.7% of findings had been dynamically reproduced and 29.4% had been reported upstream. More than $58,000 had been spent on model access. These are campaign-reported metrics, not 1,280 audited CVEs or confirmed exploits. The important development is that maintainer feedback caused the team to recalibrate severity while reproduction and disclosure remained the bottlenecks. Rob Hamilton says OpenAI's cyber-risk controls blocked deeper analysis even after he completed Trusted Access onboarding and KYC, so his team routed the work to Chinese open-source models. That is a ridiculous policy outcome: compliant American white hats get stopped while adversaries ignore the rules. BTCPay's emergency patch shows why the work matters. Machine-speed scanning has crossed into production fund protection. The ecosystem needs humans and tools that can verify, communicate, and fix what the models find.

AI Is Compressing Market Time. Bitcoin Is the Scarce Side of the Trade.

Jordi Visser argues that July's AI unwind was a speed crash, not the end of the physical buildout. His strongest point is not a bottom call. It is that AI compresses market time. Investors can discover, backtest, and crowd into the same themes faster, which creates quicker parabolas and sharper factor-level crashes even when broad indexes look calm. At the same time, compute, memory, power, and datacenter capacity remain physically constrained. Infrastructure demand can stay strong while hyperscaler shareholders face years of heavy capital spending, competition, and weaker free cash flow. Visser calls this a second printing press: governments dilute the monetary unit while AI dilutes the durability of incumbent corporate profits. That thesis is his, not a settled forecast. The bitcoin connection is still clean. In a world where both money and corporate moats are easier to manufacture, scarce bearer assets become more valuable.

Bitcoin ETFs Closed a Five-Day, $853.5 Million Inflow Streak

US spot bitcoin ETFs closed last week with five consecutive inflow sessions. The TFTC ETF tracker recorded $853.54 million of net inflows from August 3 through August 7. IBIT contributed approximately $693.64 million, or more than four-fifths of the total. Friday added $98.85 million and took total ETF assets to approximately $79.50 billion. Bitcoin still finished the session near $64,936. That contrast matters. Persistent ETF demand did not produce an immediate vertical price move, and flows alone cannot identify who sold into the bid or prove short-term price causality. They do show that regulated US vehicles absorbed nearly $854 million of net demand during a week when bitcoin remained stuck in the mid-$60,000s.

A Twelve-Year-Old JavaScript RNG Flaw Made Wallet Seeds Brute-Forceable

GitHub's reviewed advisory GHSA-rg76-677x-56q9 covers weak randomness in crypto-js versions before 4.0.0. Coinspect found that some downstream wallets used the vulnerable generator to create BIP39 recovery phrases, collapsing nominal 128-bit and 256-bit entropy into effective spaces of approximately 2^39 and 2^47 possibilities. Its monitored set contained 2,114 seeds with on-chain activity, not 2,114 confirmed victims. Coinspect measured a $5,690,922 lower-bound loss across two drain events. Merely depending on CryptoJS 3.x does not prove a wallet was exploitable. The vulnerable function had to generate the secret. Updating the app or importing the old phrase into a hardware wallet does not repair weak entropy. Affected users need a newly generated recovery phrase and must migrate funds. Never enter a seed or private key into an online checker.

⚡ FREEDOM TECH CORNER

Updating the Software Is Not the Same as Rotating the Credential

BTCPay's weekend guidance contains a broader lesson for anyone running self-hosted infrastructure. A software update and a credential rotation solve different problems.

The 2.4.2 release closes the known BTCPay vulnerability. It cannot reach back in time and erase a macaroon that may already have been copied. LND operators should regenerate macaroons, rotate other exposed authentication material, review node activity, and treat the old credentials as compromised.

This distinction applies well beyond BTCPay. Patching removes the attacker's path back into the system. Rotating credentials removes access the attacker may have carried out. If you do only the first, you may have fixed the door while leaving the stolen key active.

DATA SNAPSHOT

As of August 10, 2026, 5:05 p.m. ET

Sources: Kraken; mempool.space; TFTC Bitcoin ETF Flows.

See you tomorrow. This is not investment advice. Do your own research.

YouTube: https://www.youtube.com/@TFTC

podcast: https://www.tftc.io/tag/podcasts/