Ontology Official Announcement

🇰🇷 Ontology at Korea Blockchain Week 2026

Thank you to everyone who visited the Ontology booth in the Upbit Zone at Walkerhill, Seoul, on 30 September and 1 October, and to our Korean community for the support.

What we took away:

• The conversation has moved from trading to rails: regulation, market structure, and how everyday finance gets on chain.
• Every regulated rail needs an identity layer. Verify once, hold the result as a credential, share only the claim a service needs.
• Regulators are heading the same way. On 23 September SEC Commissioner Hester Peirce called for attribute-based verification using credentials and zero-knowledge proofs, instead of ever more data collection.

Read the full recap:
https://ont.io/news/korea-blockchain-week-2026-recap/

Coming up over the next four weeks: privacy and KYC, DID and data, and reputation.

Korean community, join us on KakaoTalk: https://open.kakao.com/o/g9RVHO8f

We are pleased to announce that the Ontology MainNet has resumed block production and returned to normal operation following an investigation into a security incident involving an ecosystem project.

Ontology’s core infrastructure remains safe, with no funds lost from the Ontology network itself. The temporary suspension was a precautionary measure taken in coordination with consensus nodes to support the investigation and mitigate potential risks.

We thank our consensus nodes, exchange partners, and community for their patience and cooperation. 🙏

Ontology has received an urgent report regarding a potential security risk involving a project built on the Ontology blockchain, which may pose risks to assets on the network.

As a precautionary measure, we have coordinated with the majority of consensus nodes, which have agreed to temporarily suspend block production to facilitate the investigation and help mitigate potential risks.

The investigation is ongoing. We will continue to cooperate with the relevant parties and closely monitor the situation. Block production will resume once the investigation has progressed sufficiently and it is deemed safe to do so.

The security of users’ assets and the integrity of the ecosystem remain our highest priorities.

We appreciate the community’s patience and understanding. 🙏

stONT service closure

Following the security and stability assessment we shared earlier today, Ontology has decided to close the stONT liquid staking service. This is a precautionary measure.

What this means for stONT users:

• The stONT service is now closed. No new staking or redemptions can be made through it.
• All ONT staked through stONT will be refunded automatically to the wallet it was staked from, within the current consensus round.
• Staking rewards from Round 289 are included in your refund.
• You do not need to take any action to receive your refund.

We will post a confirmation once all refunds are complete.

Please stay alert. The Ontology team will never message you first, ask for your seed phrase, or ask you to "claim" a refund through a link. Any message like that is a scam.

Thank you to everyone who has used stONT, and for your patience while we complete this process.

Due to a recent security and stability assessment, we have decided to close the current cycle (Round 289) earlier than originally scheduled. As a result, the current cycle will contain fewer than 60,000 blocks. This is a precautionary measure to ensure the network continues to operate within a safe and controlled environment. 🙏

🚨 Ontology Mainnet Restored

The Ontology Mainnet has successfully resumed normal operation following the completion of the required security review and network upgrade.

⚠️ All Sync Nodes are required to upgrade to v3.1.5 ASAP to ensure network compatibility and smooth synchronization.

📖 Release notes & upgrade instructions:
https://github.com/ontio/ontology/releases/tag/v3.1.5

We strongly recommend all Sync Node operators complete the upgrade ASAP and ensure their nodes are fully synchronized and operating normally.

Thank you to our validators, node operators, ecosystem partners, and community for your support and cooperation. 🙏

https://ont.io/news/ontology-mainnet-has-resumed-sync-nodes-required-to-upgrade-to-v3-1-5/

Dear Ontology Community,

We are currently coordinating with all consensus nodes to complete the required upgrade and prepare for the restoration of the Ontology Mainnet.

In parallel, all sync nodes are also required to upgrade to v3.1.4 ASAP to ensure network compatibility and smooth synchronization once the network is restored.

Please refer to the official release notes and upgrade instructions here: https://github.com/ontio/ontology/releases/tag/v3.1.4

We strongly recommend completing the upgrade as soon as possible and ensuring your nodes are fully synchronized and ready for the network restore.

We will continue to keep you updated on the restoration progress. Thank you for your support and cooperation.

Ontology Network

🚨 Ontology Mainnet Security Update

Our ongoing security review has identified malicious attack activity targeting the network.

The Mainnet will remain temporarily paused while we implement a security upgrade and conduct comprehensive verification.

✅ No user assets have been compromised or impacted.

We will continue security checks through both internal technical reviews and assessments with independent third-party security organizations to ensure the safety and integrity of the network and assets.

Our current goal is to restore Mainnet operations within 24 hours, subject to the successful completion of all required security checks and the upgrade.

Please stay vigilant when conducting transfers or transactions and always verify wallet addresses, transaction details, and links before proceeding.

🔗 Official announcement: https://ont.io/news/ontology-mainnet-pause-security-update-and-network-upgrade/

🚨 Ontology Mainnet Emergency Pause

Ontology has identified a potential security concern requiring immediate investigation.

As a precautionary measure, block production has been temporarily suspended while our technical team and validators conduct a comprehensive security review.

▶️ No confirmed security incident has been identified at this time.
▶️ There is currently NO indication of any loss or compromise of user assets.
▶️ONT, ONG, and other on-chain assets remain unaffected based on our current assessment.

The temporary suspension is a preventive security measure, rather than a response to confirmed asset loss or an ongoing attack.

We will provide further updates through our official channels.

https://ont.io/news/ontology-mainnet-emergency-pause-for-comprehensivesecurity-review/

📌 Ontology MainNet upgrade to v3.1.2

Ontology will perform a scheduled MainNet upgrade to v3.1.2 at block height 20,800,000.

What is in it

🔹 PUSH0 (EIP-3855) places zero on the stack in one byte and 2 gas, instead of two bytes and 3 gas.

🔹 BASEFEE (EIP-3198) lets a contract read the current base fee directly on-chain for 2 gas, removing the need for an external data source.

🔹 MCOPY (EIP-5656) copies memory in a single instruction. Copying 256 bytes drops from at least 96 gas to 27, which matters for encoding, byte handling and cryptographic work.

🔹 Transient storage, TSTORE and TLOAD (EIP-1153) adds low-cost state that lasts for one transaction only, at 100 gas each.

Node operators

All node operators should upgrade to v3.1.2 as soon as possible and before block 20,800,000 is reached.

Release: https://github.com/ontio/ontology/releases/tag/v3.1.2

Full breakdown 👉 https://ont.io/news/ontology-mainnet-v3-1-2-four-ethereum-opcodes-arrive-on-the-ontology-evm/

📌 Ontology at Eight: verified human data for the AI economy

Eight years ago the Ontology MainNet went live, and it has run without interruption ever since. But our eighth anniversary is not really about looking back. It is about what all of it was building toward.

AI is only as good as the data it learns from, and the industry is moving away from scraped content toward high-quality human data: consent-based, and provably created by a real person. The problem is supply. The people who create data rarely share in its value, while big tech has earned more than $1.3 trillion from user-generated data.

That is the gap we have spent eight years preparing to fill. ONTO Wallet stays a multi-chain Web3 wallet and is now adding an identity and verified human data platform: you own the data you create, build a verified profile, and earn rewards by contributing it on your own terms.

Read the full piece 👉 https://ont.io/news/ontology-verified-human-data/

📌 When human oversight becomes a compliance requirement

"We had humans in the loop" is a description of a process. "Prove it" is a demand for evidence, and evidence has properties good intentions do not. It has to name specific people, show they were distinct real humans rather than sybils or one-shot contractors, show their judgement held up over time, and make every contribution attributable, timestamped and tamper-evident.

The ground is already moving. The EU AI Act requires human oversight for high-risk AI under Article 14, and the frontier labs themselves are warning that recursive self-improvement could quietly drop the human from the loop. The moment oversight has to satisfy an auditor, attestation ("trust us, qualified humans reviewed this") is not enough. You need provenance: a record someone who does not trust you can check.

Run the self-check 👉 https://ont.io/news/human-oversight-documentation/

📌 New: when the judge shares the blind spot

Roll two fair dice. You are told at least one is a six. The probability that both are sixes is not 1 in 6, it is 1 in 11: the clue removes every outcome with no six, leaving 11 equally likely cases, one of which is the double six. The fast answer assumes an independence the clue already broke.

Avena et al. tested eight state-of-the-art models on problems built exactly to trigger that shortcut. On the counterintuitive items the models failed consistently and predictably, and chain-of-thought did not reliably rescue them. The trouble is that those same models now do the grading. When an LLM-as-judge carries a reasoning blind spot, a reward model trained on its preferences inherits it, and model-on-model agreement certifies consensus rather than correctness.

The only check that does not share the failure mode is human ground truth you can verify: evaluators whose reasoning consistency is measured and tracked over time, carried on a stable identity with signed contributions. W3C Decentralized Identifiers, W3C Verifiable Credentials, W3C Bitstring Status Lists. ONT ID and ONTO Wallet are the substrate.

Day 1 of Ontology Roundup, Issue 04.

Try the dice trap 👉 https://ont.io/news/llm-as-judge-blind-spots/

🎲 We're running a little experiment today, and we want you in it.

One dice question. One trap almost everyone falls for, the AI models included.

The poll is live on X right now 👉 https://x.com/OntologyNetwork/status/2066416847499485589?s=20

Vote, argue it out in the replies, show your working, and tag a friend who reckons they're good at probability. The more wrong answers, the better the point we're making.

No Googling. No AI. That is cheating, and it gets this one wrong anyway.

Reveal and the full piece this afternoon. 👀

A checklist today, because the SFT-vs-RL argument is eating attention that belongs one layer up.
Whichever recipe wins for reasoning models, both consume step-level human evaluation, and almost nobody can defend theirs. Five questions sort the defensible pipelines from the rest: who made each judgement, does it carry its rubric, would you notice a drifting evaluator, can experts prove credentials without exposing identity, and does revocation actually propagate.
Fewer than three yes answers means the pipeline, not the recipe, is the binding constraint.