Episode Details

James Okafor: Welcome to Pivot 5! I'm James— Maya Chen: —and I'm Maya. Let's get into it. James Okafor: Today we're covering Anthropic's Mythos model triggering cybersecurity alarms, those eye-watering $800 billion valuation talks, and OpenAI investors getting nervous. Maya Chen: Here's why this changes everything — Anthropic's latest model Mythos has government officials and bank CEOs literally losing sleep. Treasury Secretary Bessent and Fed Chair Powell held emergency meetings with major financial institutions yesterday, warning them about unprecedented cybersecurity risks. James Okafor: Yeah, and get this — the Treasury Department is now demanding direct access to test Mythos for vulnerabilities. That's unprecedented for a private AI model. Maya Chen: The data tells a different story than what Anthropic is saying publicly. While they claim Mythos has enhanced safety features, our sources indicate it demonstrated capabilities in their internal red team testing that genuinely shocked even their own security researchers. James Okafor: What nobody's talking about yet is how this connects to those recent banking system penetration tests. Three major banks failed their cyber resilience assessments last month, and now suddenly everyone's worried about an AI that could potentially exploit those exact vulnerabilities. Maya Chen: Worth noting the caveats here — we don't have specifics on what Mythos can actually do. But when both Treasury and the Fed move this fast, it suggests they've seen demonstrations that go way beyond current AI capabilities. James Okafor: The real story isn't the headline about government concern — it's that Anthropic built something so powerful they're struggling to contain it themselves. Maya Chen: Speaking of Anthropic, let's talk about those absolutely wild valuation numbers. Multiple investors are apparently throwing around offers that would value the company at $800 billion or higher. James Okafor: Honestly, I'm not buying it. That's nearly double OpenAI's already inflated $852 billion valuation, and Anthropic has a fraction of the revenue. Maya Chen: Let's look at what actually happened — these aren't signed deals. They're unsolicited offers from investors desperate to get in before Mythos launches publicly. Anthropic has resisted every single one. James Okafor: I think this is huge because it shows how disconnected AI valuations have become from any traditional metrics. We're essentially pricing these companies based on fear of missing out rather than fundamentals. Maya Chen: The data tells a different story when you dig deeper. Anthropic's revenue is estimated at $3 billion annually. An $800 billion valuation would mean investors are paying 266 times revenue. For context, that's ten times higher than the dot-com bubble peaks. James Okafor: What's fascinating is Anthropic seems to agree it's overvalued. They're turning down money! When's the last time you saw a startup do that? Maya Chen: It tracks with their cautious approach to Mythos. They know they're sitting on something game-changing and they're being deliberate about how they deploy it. James Okafor: Meanwhile, OpenAI's investors are starting to sweat about their $852 billion bet. The company's strategic pivot to enterprise customers isn't landing the way they hoped. Maya Chen: Worth noting the caveats here — OpenAI still dominates market share with 65% of enterprise AI deployments. But their growth rate has slowed from 300% to 80% year-over-year. James Okafor: The real story isn't the headline about investor concerns — it's that OpenAI is caught between two strategies. They're trying to be both a consumer ChatGPT company and an enterprise platform, and they're not executing either perfectly. Maya Chen: Let's look at what actually happened last quarter. OpenAI lost three major enterprise contracts to Anthropic, including that massive Salesforce deployment. That's what triggered this investor anxiety. James Okafor: Wow, that'