Scuttlebutt

Scuttlebutt

RBC CEO: no, I don't think we have enough guardrails on frontier AI

3 takeaways: why the US has selective access to the most advanced AI tools, how AI cut RBC's product cycles from two years to two months, and the case against mass job losses

Scuttlebutt's avatar
Scuttlebutt
Oct 21, 2026
∙ Paid

Strategic Alternatives · Oct 20, 2026 · 15m · 11m saved

With Dave McKay, President & CEO, Royal Bank of Canada

RBC CEO: no, I don't think we have enough guardrails on frontier AI

The most advanced AI safety tools are not available everywhere at once, and Dave McKay, president and CEO of Royal Bank of Canada, says that gap is a security problem, not just a commercial one. Most bank executives describe AI regulation in the abstract. Dave told the Milken Institute's Canada Dialogue, recorded for RBC's Strategic Alternatives, exactly where the access gap sits today.

"No, I don't think we have enough guardrails... these most advanced tools... are being selectively released outside of the United States." — Dave McKay

Dave runs a bank racing to patch its own vulnerabilities with the most capable models available, which makes the uneven rollout of those models his problem as much as a policy one.

The full segment is covered here so you can skip it. 15 minutes of audio, 4 minutes of reading.

Here are the 3 takeaways that matter.


This Scuttlebutt Summary is available in full to premium subscribers. If you're not one already, please consider subscribing to keep reading.


Key Takeaways

  • Dave says the US holds selective access to the most powerful frontier AI tools, which he calls a security gap, not just a commercial one

  • AI has compressed RBC's product cycles from two years to two or three months

  • Dave rejects the claim that half of white-collar jobs could vanish within a year or two

  • He expects "decomposed and recombined" value chains rather than straightforward job losses

  • RBC was "slow to capitalize" on AI deployment and is now trying to pivot

1. Selective Access, Real Risk

Dave extended Aidan Gomez's point about AI regulation into his own industry. RBC is racing to patch vulnerabilities using the most advanced models it can get, but those models are not released everywhere at the same time.

The most advanced AI safety tools are released outside the US unevenly, Dave says

"No, I don't think we have enough guardrails... these most advanced tools... are being selectively released outside of the United States." — Dave McKay

Asked directly whether the US holds an advantage in access to the most powerful tools, he did not hedge.

The US currently has selective access to the most powerful AI tools

"Currently, yes, absolutely. There's selective access to the most powerful tools." — Dave McKay

That unevenness, in his framing, is a sovereign security gap as much as a commercial one: a bank or a country without access to the best defensive tools carries more cyber risk than one with it.

2. Two Years Down to Two Months

User's avatar

Continue reading this post for free, courtesy of Scuttlebutt.

Or purchase a paid subscription.
© 2026 The Golden Age LLC · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture