# bridge exploit — X 热门讨论 (2026-09-21 11:54 UTC)
## @0xchainink (Chain INK) · 09-21 09:44 · ♥24 ↻6 💬5 $ZETA : Review 📜
What if a blockchain looked at its own numbers, decided it did not need to be a blockchain anymore, and asked its holders to vote it out of existence?
Meet ZetaChain, the universal Layer 1 that raised $27 million to connect every chain, and whose holders just voted 99.4% to shut it down and move to Solana.
What survives is ZETA itself, reborn as a Solana token, and Anuma, a private AI app the team now considers the actual product.
Let's explore one of the rarest decisions in crypto: a chain choosing to end. 👇
Marketplace Insight: ZetaChain's original pitch, a universal blockchain letting apps reach Bitcoin, Ethereum, and every other chain natively, landed in the most crowded and winner-takes-all corner of crypto, where interoperability protocols compete on network effects and second place is barely distinguishable from tenth.
The L1 never pulled enough builders to justify running its own chain. The clean contrast is that most projects in that position keep the chain alive on inertia and emissions. ZetaChain's team told holders the truth, proposed shutting it down, and 99.4% agreed. That is a failure of the original thesis and an unusually honest response to it, both at the same time.
⚪ Mission
ZetaChain's mission has shifted from connecting blockchains to connecting AI models under user control. The argument behind Anuma is that people's conversation history, preferences, and context are currently locked inside whichever AI company they happen to use, and that this memory should belong to the user, encrypted and portable across models. The token's role narrows accordingly, from gas and security for a whole chain to access for a private AI product running on Solana's infrastructure.
🔵 A Brief History
ZetaChain launched with an ambitious interoperability thesis, raising $27 million in 2023 from investors including Blockchain com and Jane Street. Its mainnet connected Bitcoin, Ethereum, and BNB Chain through an omnichain EVM that could manage native assets on connected chains without wrapping, and early on it ranked among the top blockchains by active users with over 100 million transactions and thousands of deployed contracts.
Validators included stc Bahrain, and partnerships extended to Korean telecom Stage Five and Maeil Business Newspaper's Mblock. The token told a different story. ZETA fell roughly 97% from its February 2024 high of $2.85, and monthly unlocks of around 44 million tokens kept adding supply.
The interoperability market consolidated around a handful of winners, and ZetaChain's own chain never drew the developer base its design required. In April 2026 a Gateway cross-chain exploit hit team wallets for roughly $318,000 to $334,000, with the team stating no user funds were lost, and the cross-chain deposit and messaging features at the heart of the original product were wound down afterward, with deposits closing around June. By summer the interoperability thesis was effectively already retired.
The pivot had begun in January 2026 with ZetaChain 2.0 and an AI Portal routing across multiple model providers. February brought Anuma, a private AI app built on an encrypted Private Memory Layer, which integrated a rapid run of frontier models through the spring, including GPT-5.5, Claude Opus 4.7, Kimi K2.6, and Qwen 3.6 Max. It crossed 300,000 users in mid-September.
On 17 September the team opened a 72-hour vote to shut the L1 down and migrate ZETA to Solana, citing faster execution, lower costs, deeper consumer liquidity, and existing AI-agent infrastructure, along with the ongoing cost of coordinating every upstream Cosmos security patch across independent validators. The late-August Cosmos EVM module exploits, which hit six other chains but not ZetaChain, were named as an example of that burden. The vote closed on 20 September at 99.4% in favour.
A second proposal covering the snapshot height, shutdown block, claims process, exchange conversion window, and withdrawal of assets from connected chains will follow once exchanges confirm the swap mechanism.
🔵 Ecosystem Narrative
The organizing idea has changed completely: from a chain that connects everything to an AI product that runs on someone else's chain.
➛ Anuma and the Private Memory Layer. Users' conversation history and preferences are encrypted and user-owned, following them across 35 AI models rather than staying locked inside any single AI company, with over 1.2 million requests and 30 billion tokens routed through its Bifrost gateway.
➛ A wallet in every account. Every Anuma account is a wallet, so the migration brings 300,000-plus users onto Solana on day one, into an ecosystem with sub-second execution and existing agent payment rails like x402.
➛ Direct issuance rather than a bridge. ZETA becomes a native Solana SPL token through fresh issuance, because a wrapped token would still depend on an origin chain that is scheduled to stop existing.
➛ Verifiable migration. A published snapshot export and checksum, plus a maintained archive node and explorer, let any holder independently confirm their converted balance.
➛ The Helium precedent. Helium moved its L1 to Solana across 2022 and 2023 and is generally regarded as having done so successfully, giving ZetaChain a working template for holder continuity.
➛ Frontier model routing. The AI Portal integrates models from OpenAI, Anthropic, Moonshot, and Alibaba, so Anuma sits above the model layer rather than competing with it directly, with a real-life social discovery feature now in beta.
⚪ Token Utilities
$ZETA's role is contracting from chain asset to app access token.
➛ AI credits: users lock ZETA to access AI credits inside Anuma, the primary utility after migration.
➛ Governance: holders vote through ZetaHub, including the second proposal that will finalize the migration.
➛ Staking and gas: still active on the L1 until wind-down, then retired, since Solana validators will secure native ZETA.
➛ Note on scope: ZETA on Ethereum and BNB Chain is excluded from the conversion and stays under existing terms, and balances convert from 18 to 9 decimals, rounded down.
⚪ Key Features
➛ A private AI app with user-owned, encrypted memory portable across 35 models.
➛ A 1:1 migration to Solana with no new tokens and vesting preserved.
➛ Direct issuance avoiding bridge risk entirely.
➛ Public snapshot, checksum, and archive node for balance verification.
➛ Governance that passed a chain shutdown at 58% turnout.
🔵 Meet the Team
ZetaChain is led by a compact core team that designed the universal blockchain, then proposed shutting it down when it did not find its market, which is a rarer and harder call than most teams ever make.
▶️ Core Members:
➛ Ankur Nandwani - Co-Founder | Co-founded ZetaChain and led it from the universal interoperability thesis through the 2023 raise to the decision to retire the L1 and rebuild around Anuma.
➛ Charlie Pyle - Co-Founder | Co-founded the project and has helped steer the strategic pivot from cross-chain infrastructure toward private AI, including the move to Solana.
➛ Charlie McCowan - Head of Cyber Security and DevOps | Owns security and infrastructure operations, a load-bearing role given the April gateway exploit, the ongoing cost of Cosmos patch coordination the team cited as a reason to leave, and a migration that has to move every balance without error.
➛ Jonathan Covey [ @DAOist_JC ] - Head of Community | Runs community and governance engagement, the function behind a 58% turnout on Proposal 68, exceptional participation by crypto standards.
➛ Denis Fadeev - Head of Developer Experience | Leads developer tooling and documentation, the work that will determine how smoothly builders follow ZETA and Anuma onto Solana.
➛ Backers and ZetaHub | The 2023 raise of $27 million included Blockchain com and Jane Street, and ZetaHub runs both migration votes, including the second proposal that will set the snapshot height and timeline.
🔵 Ratings
➛ Use Case: ★★★ (3/5). The original use case, universal cross-chain infrastructure, is being retired because it did not attract enough builders to sustain its own chain, and that has to be the starting point. What replaces it is genuinely interesting: Anuma's user-owned, encrypted memory that follows people across 35 AI models addresses a real problem, 300,000 reported users in seven months is fast early growth, and every account arriving on Solana as a wallet gives the token a ready-made user base. The 2-point deduction is that the new thesis is seven months old with self-reported metrics, it competes for attention against every AI interface in the market, and the token's function is narrowing from securing a chain to unlocking credits in one app.
➛ Tokenomics: ★★★ (3/5). The migration itself is handled cleanly: 1:1 conversion, no new tokens, a 2.1 billion maximum supply held constant, vesting schedules preserved, and a direct lock-for-credits sink inside Anuma. The 2-point deduction is contraction and overhang. Once the L1 shuts down, ZETA loses its gas and staking roles entirely, leaving AI credit access as the core utility. Monthly unlocks of around 44 million tokens continue, balances round down on the move from 18 to 9 decimals, and ZETA on Ethereum and BNB Chain is left outside the conversion, fragmenting the holder base.
➛ Audits: ★★★ (3/5). The migration design is genuinely security-conscious. Direct issuance removes bridge risk rather than managing it, the published snapshot and checksum make every converted balance independently verifiable, and the team cited the cost of coordinating Cosmos security patches across independent validators as a reason to leave. The 2-point deduction is the record and the disclosure. In April 2026 a Gateway cross-chain exploit drained roughly $318,000 to $334,000 from team wallets, an exploit in the core component of an interoperability chain, after which those features were wound down. No user funds were reported lost, which matters, but a demonstrated exploit in the flagship component outweighs the rest. I could not verify a named audit or security score for the Solana token or Anuma, and the migration itself is a high-risk operation not yet executed.
➛ Community: ★★★⯪ (3.5/5). 58% turnout on a governance vote is exceptional in crypto, where single-digit participation is common, and 99.4% approval for shutting down the project's own chain shows a holder base willing to accept an honest assessment rather than cling to the original story. Anuma's reported 300,000 users give the community a live product to rally around. The 1.5-point deduction is that the week's rally is governance-driven speculation on a vote rather than a completed swap, much of the original ecosystem built for the L1 will not follow it to Solana, and holders on Ethereum and BNB Chain are left navigating a separate path.
🔵 Conclusion
ZetaChain did something almost no project in this industry does. It looked at an original thesis that had not worked, a universal blockchain that never pulled enough builders to justify itself, and asked holders to shut it down. They agreed at 99.4% on 58% turnout.
What survives is a token reborn on Solana and Anuma, a private AI app with a genuinely differentiated idea: your AI memory should belong to you and follow you across every model. The risks are real. The flagship thesis failed, its cross-chain gateway was exploited in April, and the token lost roughly 97% of its value before the vote.
The replacement product is seven months old with self-reported numbers, and the token's utility shrinks from securing a chain to unlocking credits in one app. Unlocks continue, holders on Ethereum and BNB Chain sit outside the conversion, and the week's rally is speculation on a vote rather than a completed swap. The second proposal that actually executes the migration has not been scheduled.
But the bull case has a precedent. Helium made the same move to Solana and came out the other side with its network intact, and ZetaChain is executing its version with more care than most chains take over anything: direct issuance, published checksums, preserved vesting, and exchanges confirmed before a single token moves. ZetaChain spent three years trying to be the chain everything connects to.
It is now betting that owning your AI memory is the more valuable idea. The honesty of the pivot is not in question. Whether the pivot works is. https://x.com/0xchainink/status/2101970898438021432