新聞財經娛樂♬資訊 News Economy

I would like to distinguish between those vistors are A.I. or real human to access my website.

Accepted Solution

Confirm Expert: rcschroeder replied at 2025-02-09 10:47:19    

To distinguish between human visitors and AI-powered visitors, you'll need to implement a combination of techniques. Here's a suggested approach using Chrome, curl, PHP, and Python 3.9:

**1. User-Agent Detection**

Create a script that checks the User-Agent header sent by each visitor. Most browsers include their User-Agent string in the request headers. You can use a library like `php-user-agent-detector` to detect the browser type and version.

For curl, you can use the `-U` option to specify a custom User-Agent:
```bash
curl -U "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.169 Safari/537.37" https://example.com
```
**2. IP Geolocation**

Use a geolocation service to determine the visitor's IP address and location. AI-powered bots often use proxies or VPNs to mask their IP addresses, but not all of them are sophisticated enough to fake accurate geolocation data. You can use services like:

* ip-api.com (free)
* ipgeolocation.io (paid)

For PHP:
```php
$ip = $_SERVER['REMOTE_ADDR'];
$response = json_decode(file_get_contents("https://ip-api.com/json/$ip"), true);
echo $response["country"];
```
For Python 3.9:
```python
import requests

def get_geolocation(ip):
response = requests.get(f"https://ip-api.com/json/{ip}")
return response.json()["country"]

ip = requests.get("https://api.ipify.org").text
print(get_geolocation(ip))
```
**3. Browser Fingerprinting**

Create a unique fingerprint for each visitor by collecting information about their browser, operating system, screen resolution, and other characteristics. You can use libraries like:

* `php-browser-fingerprint` (PHP)
* `python-browser-fingerprint` (Python 3.9)

For PHP:
```php
$fingerprint = new BrowserFingerprint();
echo $fingerprint->getBrowserString();
```
For Python 3.9:
```python
from python_browser_fingerprint import Fingerprint

fingerprint = Fingerprint()
print(fingerprint.get_browser_string())
```
**4. Behavioral Analysis**

Monitor visitor behavior, such as:

* Click patterns (e.g., how many times they click on a button)
* Time spent on the page
* Number of form submissions
* Error rates

AI-powered bots often exhibit abnormal behavior that can be detected using behavioral analysis.

**5. Rate Limiting and Blocking**

Implement rate limiting to prevent excessive requests from a single IP address or User-Agent string. You can use libraries like `php-rate-limiter` (PHP) or `python-rate-limiter` (Python 3.9).

For PHP:
```php
$rateLimiter = new RateLimiter();
if (!$rateLimiter->isAllowed()) {
http_response_code(429);
echo "Too many requests";
}
```
For Python 3.9:
```python
from python_rate_limiter import RateLimiter

limiter = RateLimiter(max_requests=10, period=60)
if not limiter.is_allowed():
return "Too many requests", 429
```
**6. IP Blacklisting**

Maintain a blacklist of known AI-powered bot IPs or User-Agent strings. You can use services like:

* ip-blacklist.com (free)
* botscout.io (paid)

For PHP:
```php
$bot_ips = array("10.0.0.1", "8.8.8.8"); // example bot IP addresses
if in_array($_SERVER['REMOTE_ADDR'], $bot_ips) {
http_response_code(403);
echo "Bot detected";
}
```
For Python 3.9:
```python
bot_ips = ["10.0.0.1", "8.8.8.8"] # example bot IP addresses
if ip in bot_ips:
return "Bot detected", 403
```
By implementing these techniques, you can effectively distinguish between human visitors and AI-powered bots accessing your website.

**Example PHP Code**
```php
getBrowserString();
if (strpos($browser_string, 'Robot') !== false) {
http_response_code(403);
echo "Bot detected";
}

// Rate limiting
$rateLimiter = new RateLimiter();
if (!$rateLimiter->isAllowed()) {
http_response_code(429);
echo "Too many requests";
}

// IP blacklisting
$bot_ips = array("10.0.0.1", "8.8.8.8"); // example bot IP addresses
if in_array($ip, $bot_ips) {
http_response_code(403);
echo "Bot detected";
}
```
**Example Python 3.9 Code**
```python
import requests
from python_browser_fingerprint import Fingerprint

def get_geolocation(ip):
response = requests.get(f"https://ip-api.com/json/{ip}")
return response.json()["country"]

def main():
ip = requests.get("https://api.ipify.org").text
fingerprint = Fingerprint()
browser_string = fingerprint.get_browser_string()

if "Robot" in browser_string:
print("Bot detected")

geolocation = get_geolocation(ip)
if geolocation == "CN": # example country code for China
print("Bot detected")

limiter = RateLimiter(max_requests=10, period=60)
if not limiter.is_allowed():
return "Too many requests", 429

bot_ips = ["10.0.0.1", "8.8.8.8"] # example bot IP addresses
if ip in bot_ips:
print("Bot detected")

main()
```
Note that these examples are simplified and may not cover all possible cases. You should adapt and expand upon this code to fit your specific needs.