The Week AI Shook Wall Street: Opus 4.6, GPT-5.3, and a $830 Billion Wipeout

This Was the Biggest Week in AI History

If you weren't paying attention this week, you missed a seismic shift. On February 5th, both Anthropic and OpenAI dropped their most powerful AI models within minutes of each other — and the shockwaves hit Wall Street harder than anyone expected.

Here's what happened, why it matters, and what you should do about it.

The Double Drop: Opus 4.6 and GPT-5.3 Codex

Anthropic released Claude Opus 4.6 — an AI model that outperforms GPT-5.2 on knowledge work benchmarks in finance, legal, and business analysis. But the real bombshell was the announcement of Claude Cowork plugins for legal, finance, sales, and marketing — industry-specific tools that directly compete with specialized SaaS products.

OpenAI countered with GPT-5.3 Codex, their most powerful coding model ever. It's 25% faster than GPT-5.2, sets new industry highs on SWE-Bench Pro and Terminal-Bench, and — here's the wild part — it helped build itself. OpenAI's team used early versions to debug the model's own training and deployment.

The timing? Both companies had planned for a 10 AM PST release. Anthropic moved theirs up by 15 minutes to beat OpenAI to the punch.

The Market Bloodbath: $830 Billion Gone

Within hours of the announcements, software stocks went into freefall. The S&P 500 software and services index dropped nearly 4% in a single day. Thomson Reuters, LegalZoom, PayPal, Expedia, Equifax, and Intuit all crashed by double-digit percentages.

The total damage? Over $830 billion in market value wiped from software and data-centric stocks since January 28th. One single day saw $280 billion vanish.

This wasn't just an AI panic — it was a fundamental reassessment. Wall Street realized that AI isn't just making software better. AI is replacing software.

Why This Matters to You

If you work in software, legal, finance, marketing, or any knowledge work field, this week was a wake-up call. The companies that were once valued for their sticky subscriptions are now facing an existential question: can a $20/month AI subscription do what their $500/month platform does?

Increasingly, the answer is yes.

Your Move This Week

1. Try both tools. GPT-5.3 Codex is available to paid ChatGPT users. Claude Opus 4.6 is available on claude.ai. Test them on your actual work.
2. Audit your tool stack. Look at every SaaS subscription you pay for. Ask: Can Claude or GPT do this?
3. Skill up on AI agents. The era of chatbot AI is ending. The era of AI agents is here.
4. Watch the job market. The companies that crashed this week will restructure. If you're in a disrupted field, the time to pivot is now.

The Bottom Line

This week wasn't just about two AI models launching. It was the moment the market fully priced in the idea that AI is coming for the entire software stack. Not in five years. Now.

Stay sharp. Stay ahead.