{"id":108388,"date":"2026-03-04T13:00:00","date_gmt":"2026-03-04T13:00:00","guid":{"rendered":"https:\/\/www.red-gate.com\/simple-talk\/?p=108388"},"modified":"2026-02-17T10:44:11","modified_gmt":"2026-02-17T10:44:11","slug":"how-to-secure-mysql-and-postgresql-in-the-world-of-ai","status":"publish","type":"post","link":"https:\/\/www.red-gate.com\/simple-talk\/cloud\/security-and-compliance\/how-to-secure-mysql-and-postgresql-in-the-world-of-ai\/","title":{"rendered":"How to secure MySQL and PostgreSQL in the world of AI\u00a0"},"content":{"rendered":"\n

Are you feeling the impact of AI<\/a> in your daily work? No? Your database might say otherwise. If you run a forum, threads, posts and messages that are forwarded to the database are likely to be at least partially AI-generated. Run a search engine<\/a> deriving information from a database or two? You\u2019re likely to have suggestions from users or customers related to AI. AI is here and, while it\u2019s unlikely to destroy us and take over the world (at least for now), it\u2019s here to stay. That\u2019s a fact. <\/p>\n\n\n\n

Nowadays, more and more enterprises integrate AI into their applications, servers, and databases. Given the fact that databases aren\u2019t just repositories of information but backbones for your applications, securing databases like MySQL<\/a> and PostgreSQL<\/a> is becoming increasingly critical and, while we can use basic security practices<\/a> to secure our most precious data, some things have to change. <\/p>\n\n\n\n

Understanding security from the perspective of AI<\/h2>\n\n\n\n

First off, the basics of data security<\/a> remain the same: access control<\/a> is still vital, data encryption<\/a> comes without question (did you know that Google Chrome<\/a> will now throw an error if you\u2019re visiting a website not using SSL<\/a>?), regular auditing and monitoring<\/a> is still a necessity if you want to comply with privacy regulations<\/a>, and backups (you verify them, right?) are – and will – continue to be important in the age of AI.<\/p>\n\n\n\n

Considering that the evolving nature of AI introduces new challenges that can bypass conventional security mechanisms, it’s important to be aware that:<\/p>\n\n\n

\n
    \n
  1. AI-driven attackers will perform actions faster<\/strong>: while traditional security models such as web application firewalls<\/a> and signature-based malware detection are still effective if they\u2019re tested before being deployed on a live application, you need to keep in mind that nefarious parties will consult AI about your infrastructure if they\u2019re not sure. AI can (and will) make mistakes, but it will inevitably make sophisticated attacks a little faster. <\/li>\n<\/ol>\n<\/div>\n\n
    \n
      \n
    1. AI models work by aggregating public data and training on it<\/strong>: AI chatbots<\/a> like ChatGPT act like humans in conversations. To make that possible, they have to be trained on large datasets of text, and work by: <\/li>\n<\/ol>\n<\/div>\n\n
      \n