Anthropic wiped out billions in IBM market cap overnight

This past Monday, IBM's stock plunged 13%, its worst single-day decline in 26 years, after Anthropic announced that Claude Code could automate COBOL modernization, a business IBM has dominated for decades.

What is COBOL? Every time you swipe your card, withdraw cash from an ATM, or book a flight, there's a quiet piece of software called COBOL that processes those transactions in the background.

Developed in 1959, COBOL still powers the core systems of most major banks (95% of ATM transactions), airlines, insurance companies, and government agencies. As decades passed, engineers who built these systems retired, taking decades of institutional knowledge with them.  Meanwhile, replacing the architecture remains too risky and expensive to attempt.

IBM built a multi-billion dollar business around that complexity. Need to modernize your COBOL system? IBM has consultants who charge massive fees for multi-year projects. Need mainframes to process billions of transactions daily? IBM sells those too.

Anthropic's Claude Code changes the equation. The AI can map dependencies across thousands of lines of code, document workflows, and identify risks that "would take human analysts months to surface." Companies could now modernize in quarters instead of years, potentially migrating to cloud providers like AWS or Google at a fraction of IBM's cost.

Legacy code expertise has been a career moat for decades. COBOL is yet another example of how AI is systematically eliminating those moats.

Read More: CNBC

A radiologist stares at a CT scan of a patient's chest. Within seconds, AI flags a suspicious spot on the lung. Before this technology existed, finding that nodule required manually reviewing 656 separate images. Now it takes moments.

Radiology has become the fastest-moving field in healthcare for AI adoption. Of the nearly 1,000 AI devices the FDA has authorized for healthcare, roughly 75% are for medical imaging. 

The reason is urgent: radiology is facing an acute physician shortage as the workforce ages, while demand for imaging skyrockets. A single doctor can have 50 to 100 cases awaiting review daily, with some MRI studies requiring analysis of over a thousand images each.

AI doesn't replace the radiologist. It handles the exhausting needle-in-a-haystack work so the human in the loop can focus on judgment calls that actually matter: Is this spot benign, or could it be cancer? Does the patient need a biopsy? What's the treatment plan?

At IU Health's breast centers, AI now acts as a "second reader" on mammograms, catching what the doctor might miss. Adopting AI in this way means fewer missed diagnoses, faster turnaround, and doctors who can finally practice medicine instead of hunting for dots on a screen.

Read More: Indiana University School of Medicine

Subscribe here. Also, please forward to any friends who might be interested!