Crypto hacks surge in 2026 as AI cuts attack cost
Crypto hacks are surging in 2026, with incident counts at record levels and an “AI-enabled scale” pattern. By early August, CoinGecko data shows 164 separate crypto hacks, exceeding any prior full year (2025: 97). TRM Labs logged 207 hacks in H1 2026 (vs 83 in H1 2025), with Blockaid independently reporting 212. Losses in H1 2026 are estimated at about $972m–$1.1b, lower than H1 2025, partly because North Korea did not repeat the massive $1.5b Bybit hack.
The latest reporting adds a key nuance for traders: even as total losses may be lower, crypto hacks are spreading across more protocols with lower average per-incident damage. The article attributes the shift to AI, which reduces the skill floor and makes automated probing and exploitation cheaper—allowing attackers to target smaller systems.
Notable examples include BTC swap platforms Boltz and Atomiq shutting down after automated, AI-assisted probing overwhelmed small teams, and Term Labs losing about $8.5m via a governance takeover (buying/holding voting tokens and executing proposals rather than direct code-breaking). On Base, The Sandbox SAND was exploited through LayerZero-related delegate permissions, minting unbacked SAND across many transactions.
Despite headline-grabbing incidents, the piece argues smart-contract cryptography is not “collapsing.” Large losses like Drift (~$285m) and KelpDAO (~$292m) are linked more to human-layer compromise such as social engineering and multisig signer issues. For defense, AI is also being used: Anthropic’s Project Glasswing (Claude Mythos) is invite-gated for critical-infrastructure defenders to find thousands of high/critical vulnerabilities quickly.
Trading takeaway: watch for security alerts and ecosystem news around smart contracts, bridges, and LayerZero- or governance-related risk. The most immediate market reaction may increasingly center on governance and operational security headlines rather than only raw “code breaks.”
Neutral
Crypto hacks are rising because AI automation lowers attack cost and expands attacker reach. That is a persistent risk factor for DeFi, bridges, and governance-controlled protocols, which can pressure sentiment and increase intraday volatility around breach headlines. However, the reporting also suggests crypto hacks may increasingly shift from massive single-method losses toward more frequent, smaller average incidents, and key historic drivers (e.g., the absence of a repeat $1.5b Bybit hack) can keep broader drawdowns from matching worst-case scenarios. Therefore, the net price impact on any single asset is more likely to be headline-driven (short-term bearish risk around specific affected ecosystems) while the market broadly remains adaptable—leading to a neutral overall classification for price direction.