CertiK report says agentic AI is taking on crypto security and compliance work
CertiK says autonomous AI agents now investigate crypto threats, trace cross-chain funds and screen compliance in real time, with humans still accountable.
Why it matters
If security and compliance staff hand parts of their workflow to autonomous agents, firms remain legally and operationally on the hook for the agents' output, including wrong calls, prompt injection and manipulation of agents holding access to sensitive systems. The evidence covers one vendor's research rather than deployments at named crypto firms, so its demonstrated scope is limited.
Key facts
CertiK's Intel3D report says agentic AI systems now investigate threats, trace funds and act on security incidents with limited human involvement.
Reported by Crypto.News
CertiK expects AI agents to take on smart contract analysis, live transaction monitoring and cross-chain fund tracing as crypto attacks become faster and more complex.
Reported by Crypto.News
Compliance work could increasingly be handled by AI agents that screen addresses and transactions in real time and prepare regulatory reports from onchain and offchain data.
Reported by Crypto.News
CertiK described a security operations centre scenario in which an autonomous system investigates an unusual login, retrieves device fingerprint and location history, compares it against threat intelligence, decides whether to suspend an account and executes the suspension.
Reported by Crypto News (CryptoNews.net)
The report argues that agentic AI systems should be treated as workforce participants with defined responsibilities rather than conventional software tools.
Reported by Crypto News (CryptoNews.net)
CertiK said AI systems carry no legal or operational accountability, leaving deploying organisations and the humans who configure and supervise them responsible for outcomes.
Reported by Crypto.News
CertiK said each AI agent needs a defined scope of authority, escalation procedures and a human owner, and that firms treating autonomous AI as just another software tool risk shipping without adequate audit trails or behavioural monitoring.
Reported by Crypto News (CryptoNews.net)
Greater autonomy creates new risks including incorrect decisions, prompt injection and manipulation of agents permitted to interact with sensitive systems.
Reported by Crypto.News
What happened
CertiK's Intel3D report says agentic AI has moved past flagging suspicious activity into work that analysts used to do by hand across cybersecurity, anti-money-laundering and compliance. Crypto.News, citing the report, lists smart contract analysis, live transaction monitoring and cross-chain fund tracing as tasks CertiK expects AI agents to take on as attacks get faster, while compliance agents screen addresses and transactions in real time and draft regulatory filings from onchain and offchain data. Crypto News, which republished the same CertiK research, describes one example: an autonomous system in a security operations centre investigates an unusual login, pulls device fingerprint and location history, checks it against threat intelligence, decides whether to suspend the account and executes the suspension, with the reasoning recorded for later human review. The report also argues these systems should be treated as workforce participants with defined responsibilities rather than conventional software tools, which Crypto News reports as CertiK's position. Both write-ups stress that accountability does not transfer to the AI itself - the deploying organisations and the people who configure and supervise the agents stay responsible, so CertiK says each agent needs a defined scope of authority, escalation procedures, a human owner, audit trails and behavioural monitoring. No prices, funding amounts, named customers or deployment numbers appear in the evidence, and only CertiK's own research is on record here, so the scope of actual live deployments remains unquantified.
How the story developed
- First report
First report by Crypto.News
Agentic AI is becoming part of the crypto security workforce: Certik
- Update
CertiK report on agentic AI entering the crypto security workforce published
The Intel3D report describes AI agents moving beyond flagging suspicious activity into investigation, fund tracing and incident response work.
- Update
Republished account of the CertiK report
The syndicated write-up adds the security operations centre example and the claim that agentic systems should be treated as workforce participants with defined responsibilities.
How coverage built up
Independent sources over time, counted the way the consensus panel counts them: a republication of a wire story does not move the line.
- 5 Oct 2026, 13:39 UTC: Crypto.News — 1 independent source, 1 reports
Affected entities
Original sources
| Publisher | Report | Role | Published |
|---|---|---|---|
| Crypto.News Crypto media | Agentic AI is becoming part of the crypto security workforce: Certik | Original | |
| Crypto News (CryptoNews.net) Aggregator | Agentic AI is becoming part of the crypto security workforce: Certik | Syndicated |