Key Numbers

  • NEAR jumped 28% to $10.20 on May 22 (CoinGecko)
  • Weekly rally reached 45% before Friday (CoinGecko)
  • $68 million moved confidentially out of near.com in the past month (NEAR official)
  • Dynamic resharding upgrade slated for June 2026, enabling 70+ shards (NEAR press release)

Bottom Line

NEAR’s token price rose 28% after a cluster of privacy, AI, and scaling upgrades. The rally signals growing institutional interest in on‑chain AI agents and confidential finance, offering upside for holders and users of NEAR’s new services.

NEAR’s token leapt 28% on May 22 after launching confidential treasuries, AI prompt privacy, and a roadmap to 70+ shards. The move signals that investors are re‑allocating capital from Bitcoin to alts with strong AI narratives, boosting NEAR’s near‑term upside.

Why This Matters to You

If you hold NEAR, the recent upgrades raise its utility as a settlement layer for AI agents, potentially driving higher demand and price. If you trade on NEAR’s new confidential treasuries, you can now transact across 35 chains with end‑to‑end encryption, protecting sensitive business data.

Capital Rotates Into AI‑Ready Alts — NEAR Leads the Shift

Sammi Li, CEO of Ju.com, said Bitcoin’s liquidity absorption left altcoins as the next logical target. NEAR, with its long‑standing AI focus, became a “natural candidate” for investors seeking exposure to AI while avoiding Bitcoin’s volatility (Decrypt, May 22).

Li noted that momentum can outpace fundamentals, but the cautious sentiment in the market suggests the rally may be sustainable as long as the narrative holds (Decrypt, May 22).

Confidential Treasuries Open New On‑Chain Finance Pathways

Trezu’s confidential treasuries offer private multisig, payroll, and cross‑chain payments powered by NEAR’s private sharding and Intents infrastructure. In the past month, $68 million moved confidentially out of near.com, indicating early adoption (NEAR official).

These treasuries enable businesses to manage on‑chain assets without exposing balances, a feature increasingly demanded by enterprises wary of data leakage (Decrypt, May 22).

AI Prompt Privacy Boosts Enterprise Adoption

NEAR AI’s automatic PII anonymization for prompts sent to closed models (Claude, ChatGPT, Gemini) keeps sensitive data on the user’s machine, addressing inference‑cost and leakage concerns (Decrypt, May 22).

Only open‑weight models currently support full end‑to‑end encryption, but the feature positions NEAR as a preferred platform for confidential AI workloads (Decrypt, May 22).

Dynamic Resharding Will Power 70+ Shards by June 2026

Polosukhin announced that NEAR will scale to 70+ shards, surpassing Visa’s throughput, once dynamic resharding activates. The upgrade is part of network upgrade 2.13, scheduled for June 2026 (NEAR press release).

Higher shard counts will enable more AI agents to transact on‑chain simultaneously, reinforcing NEAR’s role as an AI settlement layer (Decrypt, May 22).

What to Watch

  • NEAR’s 2.13 upgrade rollout in June 2026 — could unlock 70+ shards (NEAR roadmap)
  • Q1 2026 on‑chain volume for Confidential Treasuries — expected to hit $200 M if adoption accelerates (NEAR analytics)
  • Market sentiment on alternative AI projects in May 2026 — will influence NEAR’s relative upside (CryptoPulse survey)
Bull CaseBear Case
NEAR’s AI and privacy upgrades attract institutional capital, driving price above $12 by Q3 2026 (Decrypt)Investor enthusiasm may be narrative‑driven; if fundamentals lag, the rally could reverse once Bitcoin stabilizes (Decrypt)

Will NEAR’s upcoming resharding upgrade cement its place as the go‑to settlement layer for AI agents, or will the hype fade once the upgrades roll out?

Key Terms
  • Private sharding — a method that keeps transaction data hidden from other network participants.
  • Intents infrastructure — a system that records desired actions on the blockchain without revealing details until execution.
  • Dynamic resharding — automatic splitting of blockchain shards when they reach a size threshold, increasing overall throughput.