1. Money leaves
The FBI's Internet Crime Complaint Center recorded almost $21 billion in reported losses in 2025, a record. Business email compromise accounted for more than $3 billion. Cyber-enabled fraud, where someone receives something convincing and acts on it, made up around 85% of all losses. The FBI also tracked AI as a distinct category for the first time. Reported losses are a floor. Many victims never report.
2. The channel stops working
When most inbound messages are automated, people stop reading. Real customers wait behind fake tickets. Real candidates are lost among generated applications. Real partners are ignored along with the outreach. None of this appears on a loss report, but it is often the largest cost: the channel still exists, and nobody trusts it.
3. Staff time
Someone sorts the noise. In a small business it is the owner. In a large one it is sales development, support and recruiting teams whose working day fills with triage. The cost scales with volume, and volume scales with how cheap automation gets.
4. Poisoned data
Fake leads inflate the pipeline and depress conversion rates. Fake sign-ups inflate growth. Fake reviews distort reputation. Fake comments distort policy. Every decision downstream of those numbers inherits the error.
5. Compliance and liability
Channels that accept personal data from unverified submitters create obligations: fabricated submissions often use real people's details without consent, as the New York Attorney General's investigation into fake public comments showed. Organisations that act on fraudulent requests, from account changes to data access requests, can face liability for the harm.
6. Trust
Customers who receive fraudulent messages that appear to come from you blame you. Domains without email authentication are easy to impersonate. Publishing SPF, DKIM and DMARC costs nothing and removes the easiest attack. The checklist.
Why waiting costs more
Every year the cost of producing a convincing message, application or request falls, and the volume arriving at open channels rises with it. Channels protected early keep their signal. Channels protected late have to rebuild trust with the people who stopped using them.