Cisco�s CEO says AI will be �bigger than the internet��but warns of a bubble. The money and the risks say he�s right on both.

Summary: Cisco CEO Chuck Robbins says the AI boom will create big winners and painful casualties�echoing dot-com�s lessons�while insisting the build-out will outlast any bubble. That view is reinforced and complicated by fresh data: Anthropic is reportedly raising $20B at a $350B valuation and preparing for an IPO, Nvidia is investing $2B to expand CoreWeave�s compute by 5GW despite heavy leverage, and Washington is steering toward lighter-touch AI rules via a new tech adviser. Meanwhile, Anthropic�s CEO warns catastrophic risks could emerge within years. For companies, the signal is clear: scale AI adoption, harden cyber defenses, and price in vendor and power risk.

When a veteran of the dot?com era says today�s artificial intelligence surge is a bubble – and also the next internet – it�s worth listening. Cisco Systems chair and CEO Chuck Robbins told the BBC that AI will create clear winners but �there will be carnage along the way.� He�s seen it before: Cisco was the world�s most valuable company in 2000, then lost 80% of its market value when the dot?com bubble burst. Now Cisco sells the plumbing for the AI age – and Robbins argues the build?out will endure even if some high?flyers crash.

The capital flood: real, strategic – and increasingly stretched

Consider the money. Anthropic is reportedly doubling its latest fundraising to about $20 billion at a $350 billion valuation, with backing from sovereign wealth funds and top-tier VCs and additional commitments from Microsoft and Nvidia. The company is also laying groundwork for a possible IPO this year, according to the Financial Times and TechCrunch. That�s not just exuberance; it�s a wager that demand for large models, AI agents, and safety research will keep compounding in 2026 and beyond.

On the infrastructure side, Nvidia just injected $2 billion into data center operator CoreWeave to help add more than 5 gigawatts of AI compute capacity by 2030. The deal goes beyond chips – Nvidia will help procure land and power, and its next-gen Rubin architecture and storage/CPU lines will be baked into these �AI factories.� Yet the balance sheet tells another story: CoreWeave had $18.81 billion in debt as of September 2025 against $1.36 billion in Q3 2025 revenue. That�s a classic bubble tell – heavy leverage chasing explosive demand – and a reminder of Robbins� point: winners emerge, but not everyone survives.

Speed vs. safety: a widening gap

Robbins warns AI will supercharge cyberattacks and scams, making phishing and impersonation more convincing. He says the industry will counter with new defenses, including quantum-era approaches. But on existential risks, others are louder. Anthropic CEO Dario Amodei published a nearly 20,000?word essay arguing the world could see catastrophic misuse within a few years – from bioterror to authoritarian surveillance – and that current guardrails are not enough. His line is stark: humanity is being handed �almost unimaginable power,� but society may lack the maturity to wield it.

That tension is the industry in microcosm: the same company reportedly raising $20 billion and prepping for the public markets is also warning policymakers to move faster on safety. The UK�s 2023 Bletchley Park summit put a marker down globally, but the regulatory crackle is now in Washington.

Policy winds: lighter touch in Washington, shifting chips in geopolitics

Robbins, who also chairs the Business Roundtable, says the Trump White House has been unusually accessible to business leaders. At the same time, Silicon Valley has a new conduit into the West Wing. The FT reports that Sriram Krishnan, a former engineer and VC, is now a key AI adviser shaping a light?touch federal approach – drafting executive orders to preempt stricter state rules and reportedly helping soften export controls so older chips can still flow to China. The pitch to industry is clear: move fast, build, and keep America (and its partners) ahead in the AI race.

Supporters call Krishnan the �connective tissue between Silicon Valley and Washington.� Critics worry that easing constraints without strong safety baselines amplifies systemic risk – exactly the concern Amodei raises. Markets, for now, are voting for speed: capital is green?lighting compute, data centers, and model training at historic scale.

Jobs, operations, and the CIO�s risk ledger

Robbins is blunt about the workforce: some roles, especially in customer service, will shrink; others will be reshaped. His advice is practical: don�t fear AI – master it, or risk being outcompeted by someone who does. For operators, the playbook is getting clearer:

  • CIOs/CTOs: Stress?test vendor concentration and power availability. Contracts should hedge for capacity, price volatility, and model switching.
  • CFOs: Watch leverage and circular deals across the stack. If your provider is funding capex with aggressive debt, your uptime and pricing risk go up.
  • CISOs: Prepare for rapid growth in high?fidelity phishing, voice cloning, and automated intrusion. Raise authentication standards and invest in anomaly detection.
  • Ops/Facilities: Land and power procurement is now strategic. Expect long lead times, permitting friction, and rising interconnection costs.
  • HR/L&D: Roll out AI upskilling tied to explicit KPIs; �pilot purgatory� is expensive without measurable productivity gains.

Dot?com echoes without the same ending

Jamie Dimon says some AI money will �probably be lost.� Sundar Pichai cautions there�s �some irrationality� in the boom. Robbins agrees – and still insists AI will be bigger than the internet. Looking at the numbers, both halves can be true: towering valuations, leverage, and policy tailwinds can inflate a bubble. But unlike 2000, the build?out isn�t vapor. It�s steel, power, and chips – hundreds of billions flowing into infrastructure that businesses will use, even if ownership changes hands during the shakeout.

The question for leaders isn�t whether the AI cycle ends. It�s whether your organization is structured to capture the upside – and withstand the whiplash – when it does.

Found this article insightful? Share it and spark a discussion that matters!

Latest Articles