Good Ai vs BAD Ai (Round 1)

What happens when both the good guys and the bad guys have a magic power? One that allows them both to see through walls, around corners, and move at the speed of light.

Stalemate?

Only if both sides are ready for the fight.

In late August, several AI agents discovered a PHP vulnerability that impacted over 70% of websites worldwide. The issue was caused by the way PHP dealt with running out of memory, leaving sensitive information like passwords and account information vulnerable to bad actors.

On October 21st, two months after the security risk was identified, PHP released a patch that closed the vulnerability. For the five servers we run at FRANK Data Insights, the update took about 5 minutes per server and didn’t create any problems with our applications. But the long-term impact of all of these vulnerable sites may not be realized for months and years to come.

Round 1 (TKO Good AI)

For a small business dealing with the new superpowers that AI gives the bad guys, an issue like this can be debilitating. Understanding the risk and managing the triage requires expertise that most small businesses don’t have, not to mention the exposure that enterprise environments faced for two months.

In later rounds of this fight, Good AI was able to assist with the fix, but the damage in the early round was done and cannot be ignored.

A five-minute fix. A two-month opening.

So what are the solutions for these issues that are sure to grow as AI agents’ abilities continue to advance at an unbelievable rate?

Solutions

For those who manage their own environments:

1. Know your server inventory and what runs on each.

At the small and medium business level, the focus of operations is efficiency and getting the work done. Security is important, but in today’s world, security must become a priority for all businesses, even the small ones.

2. Scan your servers.

Daily scans on the OS, weekly alerts on medium and low vulnerabilities, with immediate alerts on critical risks. The same way you pay attention to your costs and ROI, prioritize security across your POS, websites, and any other system used to get work done.

3. Use tools available to your environment.

If you run in AWS, use GuardDuty to help identify and triage any issues. Azure has Windows Defender that does similar scans depending on the environment.

4. Become familiar with AI tools.

Fight fire with fire. Many AI models can scan your current stack and give you a baseline of security issues. If the bad guys are using AI to discover your security issues, you should use it to defend against security issues.

5. Reach out to us at FRANK Data Insights.

If all of this sounds like Greek to you, don’t worry. Reach out to us at info@frankdatainsights.com. We will give you a free assessment of your current security stack and offer a plan to keep you informed and your environment secure.

Don’t wait for the next round to get in your corner.