A Bitcoin Improvement Proposal for the post-quantum signature scheme SHRINCS has been published, putting a concrete Bitcoin-focused option into the debate over how the network could defend itself against future quantum attacks. The proposal comes from research led by Blockstream, whose team has already tested SHRINCS in production on the Liquid sidechain. Blockstream Research’s Jonas Nick described it as the first concrete post-quantum signature proposal designed specifically for Bitcoin, while also stressing that it is not meant to be Bitcoin’s final signature system and is not optimal in every respect. The core appeal is size. Existing post-quantum schemes endorsed by the National Institute of Standards and Technology are far larger than Bitcoin’s current ECDSA and Schnorr signatures, creating a major throughput problem if deployed directly on-chain. SHRINCS is still much bigger than Schnorr, but materially smaller than many alternatives, and Blockstream’s estimates suggest Bitcoin could still run at roughly 3 transactions per second under SHRINCS, versus 0.5 TPS for ML-DSA and 0.36 TPS for SPHINCS+. The catch is that SHRINCS is still early. Its BIP says the security proof remains unfinished, and the design introduces statefulness, signature growth over time, and a large fallback recovery path if a signing device is lost. Blockstream is also exploring lattice-based signatures and zero-knowledge proof aggregation, which it estimates could lift Bitcoin throughput to 6.7 TPS if combined with SHRINCS.

Fastino has released GLiNER2.5, a new information extraction model built on a boundary prediction architecture rather than the traditional span enumeration approach. The model directly predicts where entities begin and end, removing the maximum entity length constraint and allowing it to process contexts of up to 4,096 words. According to the report cited by Techub News, GLiNER2.5 also keeps computational complexity linear with sequence length and supports joint decoding of entities and relations. On benchmark performance, GLiNER2.5 posted a macro F1 score of 56.17 across 16 zero-shot benchmarks, slightly ahead of its predecessor, GLiNER2. Fastino has released three checkpoints on Hugging Face with 74 million, 194 million, and 287 million parameters under the Apache 2.0 open-source license. The model is aimed at use cases in legal contracts, medical documents, financial services, insurance claims, and AI safety tools. Tasks listed in the report include personally identifiable information detection and redaction, contract clause extraction, and knowledge graph construction. Fastino said the model supports local inference deployment on CPU, CUDA, and MPS environments.

Cybersecurity firm JUMPSEC said the North Korea-linked hacking group BlueNoroff has been using fake Zoom and Microsoft Teams meetings to scan cryptocurrency wallets and spread malware through compromised Telegram accounts. According to JUMPSEC, the phishing kit works on both Windows and macOS and is capable of stealing browser keys. The attackers also used AI-generated profile images to fake meeting videos, then scanned wallet extensions to identify high-value targets before delivering malware. JUMPSEC added that security researchers were able to recover the source code after the attackers exposed a JavaScript map. The disclosure was cited by crypto.news and later carried by Techub.

Prosecutors in São Paulo state have filed a public civil lawsuit against Tools for Humanity Corporation, the developer tied to World Network, and Amazon AWS Serviços Brasil, alleging the companies illegally collected iris biometric data from more than 400,000 people in exchange for financial compensation. The São Paulo Public Prosecutor’s Office is seeking at least R$240 million in collective moral damages, roughly $47 million, and is asking the court to permanently ban the practice of offering money for iris data. Prosecutor Patrícia Leitão also requested an injunction to preserve all records, metadata, contracts, and technical reports tied to the handling of the biometric information, while seeking clarity on which Amazon group entity actually hosted the data. The case builds on an earlier order from Brazil’s National Data Protection Authority, which on Jan. 24, 2025 told Tools for Humanity to stop offering cryptocurrency or other compensation for Brazilians’ iris scans, and on a 161-page São Paulo city council report that cited serious violations across data protection, consumer protection, and financial regulation.

Coinbase now supports direct Indian Rupee (INR) transfers via IMPS, letting users deposit and withdraw without P2P risks. Spot, perpetual futures, and local order books are live. Rollout is gradual.

Coinbase has added direct INR deposits and withdrawals in India through IMPS, giving users a bank-linked way to access spot trading, perpetual futures, and local INR order books.

Circle Payments Network adds payout corridors in India, Singapore, Philippines, UAE, EU, and US. Near-instant settlement via domestic rails like SEPA, FAST, Fedwire. USDC stablecoin backs compliant cross-border flows.

Coinbase rolls out direct INR transfers in India, enabling IMPS-based instant deposits and withdrawals, alongside spot and perpetual trading with local order books, after FIU-IND registration.