What happens when decades-old cons meet modern automation? The answer, according to recent reporting, is a steep and record-breaking rise in losses.
Bots and AI: Old scams, new scale
The Register reports that "Bots are now firmly in the toolbox, helping crooks scale old scams" and that "Crims are taking advantage of AI to sharpen old scams." Those two short observations capture a key shift: the tools of automation and generative models are being applied not to novel schemes but to established frauds, amplifying reach and efficiency.
FBI tallies a record year
The FBI reported Monday that cybercrime losses hit a record $20.87 billion in 2025, with help from bots. That single figure, pushed to an all-time high, is the anchor for understanding why the automation of fraud matters beyond anecdotes and technical demonstrations.
Why this matters — perspectives to consider
- For technologists: The combination of AI and bots rewrites the economics of fraud. If automation lowers the time and cost to target more victims or tailor lures at scale, detection and mitigation must become both faster and more automated to keep pace.
- For policymakers and law enforcement: A record loss figure tied explicitly to bot-enabled activity signals a changing threat landscape. Resource allocation, legal frameworks, and cross-sector coordination will all be measured against how quickly they adapt to an environment where automation multiplies old harms.
- For users and organizations: The greatest risk may be familiarity. Sophisticated automation can make longstanding scams—already familiar to many—appear more credible or omnipresent. Defensive steps that once sufficed against low-volume fraud may prove inadequate when attacks arrive at machine speed.
- For adversaries: The appeal is obvious: reuse known social-engineering playbooks while deploying bots and AI to boost outputs. The result is not necessarily novel deception techniques but much larger scale and efficiency.
Looking ahead
The Register’s reporting and the FBI’s loss estimate together frame a clear dilemma: reducing harm requires systems that are as scalable as the threats. That may mean more automation in defenses, different legal and investigative tools, and wider public education focused on scaled deception rather than just novel tricks. If the past year shows anything, it is that making scams smarter with AI is less the problem than making them faster and larger with bots.
How quickly defenses and institutions adapt will determine whether the next headline reports incremental loss or a return from record highs.
