AI can not only pose information security risks but can also seriously damage your company’s reputation. For my first AI ISO 42001 training, I developed three major categories of risks for companies where employees are using AI at different levels and with various tools:
1. Quality risks
2. Plagiarism
3. Information security risks
We’ve already discussed AI’s information security risks in recent posts and articles. But where can plagiarism come into play? If an AI system is still new and hasn’t been trained on enough data in a specific field, it may adopt large chunks of existing data to generate new content. Some of this data may be plagiarized or owned by other companies.
Quality risks can arise from AI’s ability to mimic the creative work of humans without fully understanding the context. Have you ever seen nicely worded texts in newspapers or on LinkedIn where, despite the correct use of words, the overall text doesn’t make sense or lacks any significant or valuable insights? This is called “AI hallucination,” where the AI tries to generate content but, due to insufficient data or unclear prompts, creates something that only resembles what it thinks a human would produce.
Failing to review the final output from AI can lead to reputational damage. Whether through plagiarism or delivering a product (text, drawings, or software) that mimics something meaningful but doesn’t actually make sense, the consequences can be severe for your company’s reputation.
