Changpeng Zhao pushed back in late January and early February against claims that Binance engineered the October 10, 2025 flash crash, describing the narrative as “imaginative FUD” and a “coordinated attack.” Analysts cited in the text pegged the episode at roughly $19–$28 billion in deleveraging and about $1.75 billion in liquidations, turning what looked like a sudden price event into a broader stress test for market structure and liquidity.
Zhao’s core message was that there was no orchestrated $1 billion Bitcoin dump, and that visible wallet swings reflected user withdrawals rather than Binance shifting its own inventory. He paired that denial with public explanations and compensation-related messaging, leaning on the idea that the most visible “signals” were being misread in the middle of a fast-moving unwind.
— CZ 🔶 BNB (@cz_binance) February 2, 2026
What the post-crash response actually looked like
The remediation package described in the reporting was designed to restore confidence quickly, but it also highlighted how much of the recovery depended on discretionary, centralized actions. The text says Binance allocated $1 billion from its Secure Asset Fund for Users to Bitcoin as a stabilization measure, paid roughly $600 million to customers impacted by technical glitches, and later communicated a $400 million recovery plan plus a $45 million airdrop as part of the broader reset.
Even with those measures in motion, the event exposed uncomfortable frictions in liquidity, conditional market making, and platform reliability that can’t be solved with compensation alone. In other words, the response may have repaired balances for many users, but it did not eliminate the underlying question institutional desks care about most: how the venue behaves when the system is under maximum load.
From an execution and UX standpoint, the most damaging failure mode was the mix of access issues and oracle-driven pricing distortions at the exact moment users needed control. The text points to an internal oracle problem that contributed to USDe’s depegging, along with reports of users being locked out or unable to manage positions during rapid moves, and it cites an estimate of roughly 1.6 million accounts being “zeroed out” as the situation cascaded.
Zhao tried to reposition the conversation away from blame and toward “systemic fixes,” while external voices pulled in the opposite direction by demanding more transparency and formal scrutiny. He called the allegations “far-fetched” and spoke “as a shareholder and user rather than an active executive,” while Star Xu and Cathie Wood publicly pushed back and called for greater transparency, and the text notes that former regulators urged a formal investigation.
What counterparties will pressure-test next
For institutional treasuries and professional traders, the operational takeaway is that venue risk is not just about solvency—it’s about whether you can execute, exit, and reconcile positions when the interface and pricing stack are under stress. That reality will drive tighter exchange selection processes, higher margin buffers, and more conservative automated unwind thresholds, especially for participants who can’t tolerate being trapped in a volatility spike.
Looking forward, the competitive bar is going to be set by exchanges that can prove—rather than promise—that transaction state is clear, oracle feeds are hardened, and “steps per operation” don’t explode during a crisis. The text also flags that ongoing negotiations over Binance’s external monitor will remain part of how counterparties model governance and oversight risk, which means the reputational and supervisory dimension will sit alongside pure execution metrics in venue selection.
