Nearly half of AI generated code may have security flaws

Studies show that almost 50% of code produced by AI tools can contain vulnerabilities, even from large language models. This raises concerns about the safety of relying solely on AI for coding critical software 🐞🖥️.