What happened
CertiK, one of the largest smart-contract audit and on-chain security firms in crypto, said agentic AI has moved from pilot into live security work, per a Crypto. News report dated October 5. The firm's framing, as summarised in that report, is that autonomous systems are now investigating threats, tracing funds across chains, and acting on security incidents with limited human intervention.
The verbs matter. Investigating, tracing, acting. Those are the three jobs a tier-two SOC analyst or a blockchain forensics associate gets paid to do.
CertiK is saying software is doing them now, inside the same workflow, not in a lab. The scope the firm cited stretches beyond threat hunting into anti money laundering and compliance, which is where the regulatory stakes start to bite. AML queues at exchanges and custodians are the single largest line of security headcount in the industry.
If agentic systems are already triaging alerts, pulling address clusters and preparing filings, the labour math on a crypto compliance desk changes fast.
Why it matters
The crypto security workforce has spent two years losing ground to attackers. Exploit losses crossed $2. 2 billion in the first nine months of 2025, per CertiK's own earlier Hack3d reporting cited by Crypto.
News, and laundering flows through cross-chain bridges and mixers routinely clear hundreds of millions in a single incident. Human teams can't read mempool, scrape socials, cluster addresses and draft an incident note in the ninety seconds it takes a draining script to route funds through three chains. Agentic AI, as CertiK describes it, is pitched at exactly that gap.