Prompt
How do I catch bot traffic that is creating fake accounts?
Latest observation
Oct 29, 2026
OpenAI APIWeb search: off
To catch bot traffic creating fake accounts, use a layered approach: detect, delay, challenge, and verify. The goal is to identify suspicious signups without hurting legitimate users too much.
1) Add friction to signup
- CAPTCHA / reCAPTCHA / hCaptcha on signup and suspicious flows.
- Email verification before account activation.
- Phone verification only if appropriate for your product.
- Rate limits on signup attempts by IP, subnet, ASN, device, and email domain.
2) Track signals that bots often reveal
Log these on every signup attempt:
- IP address, ASN, geolocation
- User agent
- Device/browser fingerprint signals
- Request velocity and repetition
- Time to complete form
- Mouse/keyboard interaction patterns
- Disposable email domain usage
- Failed verification attempts
- Reuse of names, usernames, or referral codes
- Proxy/VPN/Tor indicators
3) Build bot-detection rules
Examples:
- Too many signups from one IP in a short time
- Many accounts from the same /24 or ASN
- Same device fingerprint used across many accounts
- Signups completed unrealistically fast
- High ratio of disposable or newly created email domains
- Multiple accounts with similar usernames or profiles
- Mismatch between IP geolocation and claimed region
- Traffic from known hosting providers, proxies, or datacenters
- Repeated use of the same referral/discount/promo code
4) Use risk scoring
Assign a risk score to each signup based on signals:
- Low risk: allow normally
- Medium risk: require email verification or CAPTCHA
- High risk: block, queue for review, or require stronger verification
This is usually better than a single hard rule.
5) Watch post-signup behavior
Bots often look normal at signup but act oddly later:
- Very fast profile completion
- Immediate bulk actions
- Copy-pasted bios or messages
- Repetitive navigation patterns
- Many accounts linking to the same external URLs
- Low engagement or very short sessions
6) Use honeypots
Add hidden fields that real users won’t fill in, but bots may:
- Hidden form inputs
- Invisible links/buttons
- Form fields that should remain empty
If they’re filled, flag the signup.
7) Protect against automation
- Check for headless browser indicators
- Detect abnormal JavaScript execution patterns
- Use device/browser attestation where possible
- Block or challenge suspicious automation frameworks
8) Review and iterate
- Keep a list of confirmed fake accounts and their signals
- Feed those patterns back into rules
- Measure false positives so you don’t block real users
- Update thresholds as attackers adapt
9) Good tooling options
Depending on your stack, you can use:
- CAPTCHA providers: hCaptcha, reCAPTCHA, Turnstile
- Bot management/WAF tools: Cloudflare Bot Management, Akamai, Fastly
- Fraud/risk platforms: Sift, Arkose Labs, DataDome
- Custom analytics: event logging + rules engine + scoring
10) A practical baseline setup
If you want a simple starting point:
- Require email verification
- Add CAPTCHA only when risk is high
- Rate-limit by IP and ASN
- Block disposable email domains
- Log device/IP/UA/fingerprint
- Score signups and review suspicious clusters
- Monitor post-signup behavior for abuse
If you want, I can help you design a concrete detection plan for your stack, like:
- Node/Express
- Python/Django
- PHP/Laravel
- Cloudflare
- AWS
- A custom risk-scoring system