Predictive text is software that suggests or completes words as a user types. It can speed up writing, reduce errors, and make mobile or office applications easier to use. Common examples include word completion, phrase suggestions, and automatic corrections that appear before a word is fully entered.
In cyber security, predictive text matters because it can influence what data is displayed, stored, or sent to a cloud service. If suggestions are generated using remote models, typed content may cross a trust boundary and raise privacy or compliance concerns. Administrators may disable it to minimize data exposure, enforce policy, or keep sensitive drafting inside a controlled environment. Attackers can also abuse autocomplete-style features to trick users into accepting the wrong text, especially in phishing, invoice fraud, or code injection scenarios. For defenders, the key question is not convenience alone, but whether the feature fits the organization’s data-handling rules.