Data poisoning is a term used in the fields of artificial intelligence, cybercrime and cybersecurity as well as big data and smart data. It describes the targeted manipulation of data that machines and algorithms use for learning. The aim of data poisoning is to falsify the quality or results of these systems and thus bring about unwanted or even harmful decisions.
Imagine a company uses artificial intelligence to sort applications. If someone deliberately feeds false or misleading data into the system - for example, manipulated CVs - this can lead to unsuitable candidates being selected later as particularly suitable. The system is then "poisoned" and no longer makes reliable decisions.
Data poisoning is a major challenge for companies that use technology such as AI or big data. Companies that protect their data quality and strictly regulate access rights can better protect themselves against such attacks. Awareness within the company is therefore particularly important to ensure that potential vulnerabilities are quickly recognised and closed.