Pinned Tweet
Resolving the tension between security and innovation is simple, but not easy: We need visibility and control across every system that agents touch.
If this holds up, wow. So much of modern security depends on cryptography, which depends on math. AI could accelerate discoveries that change what we can trust. Updating cryptography across millions of connected systems is an industry-wide challenge. We need to be ready.
“When the race to the top is working properly... everybody wins.” Well said. Reminds me of a great line in Anthropic’s IPO letter: “If our competitors adopt our practices and improve on them, we consider that a success.” That’s the kind of competition I’d like to see everywhere. Tech could use more of it. Cybersecurity especially. Congrats to everyone at @AnthropicAI!
Remarks from Anthropic CEO @DarioAmodei on our initial public offering:
My co-founders and I started Anthropic because we believe that the most powerful technology of our time should be built for the benefit of humankind.
Daniela and I lost our father to a disease that, just a few years later, became curable. I started my career in biology, hoping to help future cures come sooner. In the lab, I saw the vast thicket of complexity that scientists face when they’re trying to understand disease and develop new treatments.
So many of the scientific tools that can cut through that thicket—tools for gene editing, gene synthesis, controlling brain cells with light—started as pure ideas. In other words, they came from thinking. From intelligence.
Human intelligence has already made diseases like plague, smallpox, and polio history for most of the world. But progress on other diseases has been too slow for so many patients and their families. Going from a great idea to a treatment can take decades: it took many years from the first discovery of a strange repeating pattern in DNA to the approval of the first treatment for a disease using CRISPR.
Nature is full of these patterns. But we’ve had too little intelligence to decipher most of them; too few scientists, with too little time. Now we can build that intelligence. Whether it’s protein folding, drug design, cancer genetics, or clinical trial administration, scientists across many areas of biology and medicine are finding Claude can help them do their research in a fraction of the time it would’ve taken just a couple of years ago. I believe sufficiently powerful AI, working in concert with humans, could compress a century of scientific and medical progress into a decade.
I don’t want to get ahead of myself. At this point all we have are early results. Experiments and clinical trials will still take time. And although they involve slow and ponderous processes, I believe those processes can be substantially accelerated by AI. We hope that in the future fewer families miss a cure by just a few years.
Medicine is just a particularly vivid example, but there are many others. Our customers—the world’s most ambitious companies, nonprofits, and individual people—are already applying the intelligence we’ve built to stimulate economic growth; lift people out of poverty; protect democracy and free societies; even make everyday services like the DMV work better. We’re proud to work alongside them. And I’m extremely optimistic about what we can do together in the future.
At the same time, we have always been honest that this technology has very real risks. These range from misuse of the technology by individuals or nation-states, to economic disruption, to losing control of AI systems. We discuss these risks not because we think they are inevitable, but because we think they are avoidable.
Avoiding them depends on the choices made by the people building AI. There is a gigantic difference between how AI will go if we pay little attention to safety, versus how it will go if we are obsessed with safety. And an obsession with safety and with public benefit has been woven into Anthropic’s DNA from the start. From safeguards against biological and cyber risks, to alignment training and testing, to looking inside our models like an X-ray, Anthropic has pioneered methods for understanding and steering the behavior of AI systems.
There is much work still to be done, and the most difficult safety work lies ahead of us. The incredible talent of the company that has enabled us to build the most capable AI systems will need to be directed, even more so than it has already, towards double- and triple-checking that those AI systems are safe and secure.
And to state it unequivocally, we will pace the training and deployment of AI systems, taking as long as is necessary to have confidence that they are safe and will remain under human control. We will not bend to either business or political pressure. Today puts the company in a stronger financial position, and in doing so only strengthens our resolve to do the right thing.
If we can show it’s possible to make the right choices on safety and still build an incredibly successful company, others in our industry will be inspired to do the same. We call this a race to the top. When the race to the top is working properly, it doesn’t matter which company has the best models—everyone wins because everyone is building safely. The race to the top works best when driven from the front of the pack, and we fully hope and expect to do so, but the goal is to make the whole industry better, regardless of whether Anthropic is winning or losing at any given time.
The advent of machine intelligence is one of the seminal moments in human history, and we have a rare opportunity to shape it for the better.
This moment is bigger than any one company, or any one country. It will take all of us, working together, to ensure this technology goes well for us and the many generations to come. Starting today, we hope that you’ll join us on this mission.
When we say independent and neutral, we mean it. Okta for AI Agents now supports third-party, XAA-enabled identity providers. okta.com/en-sg/newsroom/pres…
I’d also ask what protects my AI workload between the day a flaw turns up and the day it’s patched. If one account gets compromised in that window, what else can it reach? You need multiple defensive layers in place.
“When we do look around corners, it's like a zombie apocalypse headed our way right now.”
- @CoreWeave CISO Jim Higgins describing the cybersecurity threat landscape facing AI infrastructure at today's Aspen Cyber Summit in DC.
Higgins also warned against putting critical AI workloads in shared data centers and expressed concern about drone attacks and physical threats to data centers.
@AspenDigital
With cyberattacks in the headlines again, one message bears repeating: the more technology can do, the more security fundamentals matter.
Phishing-resistant MFA for people. Least privilege for every identity. And as businesses deploy AI agents, an identity, an owner, and clear limits on what each agent can access and do. Getting these fundamentals right takes sustained work. They’re essential to adopting new technology with confidence.
Appreciate the security teams doing that work every day.
Workforce productivity is just the start. The bigger opportunity is using agents to transform customer experiences and grow your business. Here's how we’re helping brands pursue that opportunity securely with Auth for UCP: linkedin.com/pulse/unlocking…
5
1,037
Phishing resistance isn’t the flashiest topic in AI security, but it’s one of the single most effective ways companies can protect themselves right now.
Always great to join @jimcramer to cut straight to what matters. Following the critical PHP vulnerabilities uncovered by frontier AI models, there’s a lot of fear about how fast threat actors can find flaws and chain attacks together. That threat is real, but expecting enterprises to patch decades-old software overnight isn’t realistic on its own. That’s why layered defense and modern identity are so critical.
The fact is that far more organizations will be hacked today through phishing than a zero-day software vulnerability. With AI supercharging social engineering, punishing employees for clicking a convincing link is a losing game. We already have the tech to stop the root cause: When Okta researchers analyzed 6,000 real-world social engineering attacks this month, strictly enforcing phishing resistance at sign-in would have stopped 99.75% of account takeovers.
At the same time, identity is now about more than just humans. As we’ve seen with major banks restricting consumer AI agents from accessing customer accounts, companies urgently need standards to distinguish trusted agents from malicious ones.
Securing this era, from human sign-ins to the new wave of agents, requires the entire security ecosystem to step up, drop the complexity, and work together through efforts like @the_blueprintai. Customers shouldn’t have to be configuration experts to stay protected against preventable attacks.
Watch the full conversation below.
I recently spoke with @ErleNorton at @WSJ to talk about leadership and staying challenge-driven.
When you lead a company from day one to $3B+ in revenue, every milestone presents a totally different challenge. Early on, it was simply proving we could build something people wanted. Today, leading an entire company through the agentic AI platform shift requires a whole different muscle: conditioning the organization to embrace change.
As leaders, especially during tech shifts, it’s tempting to look for an expert playbook. But the reality is that no one has all the answers. You have to trust your instincts, get out of the building to listen to customers, and give your teams a reason to care about where you’re going.
Check out the full conversation here: wsj.com/articles/okta-ceos-c…
5
452
It was great to spend last week visiting with employees, customers, and partners across Germany, Switzerland, and the Netherlands.
I saw firsthand how European companies are moving every bit as fast on AI agents as companies in the US, but hitting the exact same challenge: The broad access that makes agents more capable and valuable are the very things that introduce new security risks.
To innovate without sacrificing security, organizations need visibility and control over their agentic stack. They need to know where agents are, what they can do, what they’re actually doing, and how to respond if something goes wrong.
Right now, every vendor in the market is claiming they can do it all, leaving enterprise customers confused about how to actually secure their AI roadmaps. Our goal is to bring clarity to that conversation: explaining how @Okta provides the necessary identity foundation in collaboration with the broader ecosystem through the @the_blueprintai.
We closed out the trip by joining a town hall at our office in Amsterdam. EMEA is a massive driver for Okta, and our local teams are executing with incredible focus to help our customers succeed. What we offer customers has never been more relevant and the opportunity ahead of us in the region is massive.
Thanks to everyone across EMEA who made this trip so impactful. Energized for what's ahead!
Great to see the momentum behind Personal Agent Protocol, and proud that @Okta is helping shape it.
On Tuesday, Sierra announced Personal Agent Protocol with Meta, Genesys, Instinct, Rocket, Shopify, Stripe, and Walmart. The response has been incredible and today we’re publishing the first draft of the protocol at personalagentprotocol.org and announcing 35 new design partners.
As personal agents take on more, our goal is simple: consumers have choice and control, and companies know when they’re dealing with an agent and who it represents.
This draft is a starting point. We’d love for companies, developers, and personal agent builders to read it, give feedback, and help us build an open standard together.
Todd McKinnon retweeted
On Tuesday, Sierra announced Personal Agent Protocol with Meta, Genesys, Instinct, Rocket, Shopify, Stripe, and Walmart. The response has been incredible and today we’re publishing the first draft of the protocol at personalagentprotocol.org and announcing 35 new design partners.
As personal agents take on more, our goal is simple: consumers have choice and control, and companies know when they’re dealing with an agent and who it represents.
This draft is a starting point. We’d love for companies, developers, and personal agent builders to read it, give feedback, and help us build an open standard together.
The secure agentic future needs open standards. Cross App Access is an important step toward giving enterprises a consistent, secure way to manage how agents connect across applications, and toward making solutions like Okta’s Agent SSO possible across the ecosystem.
Really enjoyed talking with @dsp_ at Oktane about the role that open standards like MCP and Cross App Access will play in building a secure, interoperable agentic future, and about the work @AnthropicAI, @Okta, and others across the ecosystem are doing to make that future real.
Thanks David for the great discussion and for helping push this work forward.
2
11
1,212
As AI agents move from experiments into real work, every company is figuring out what this platform shift means for their business. @Dell has been one of the customers pushing us forward—challenging us, helping evolve our thinking, and keeping us focused on what enterprises actually need.
I was grateful to have Doug Schmitt join me at Oktane to share how Dell is approaching this shift at global scale.
Thank you, Doug, for your trust, and for continuing to push us forward.
5
15
1,430
Todd McKinnon retweeted
Despite a few travel surprises, great to join @toddmckinnon on stage at the Oktane keynote yesterday.
As enterprises scale AI agents, they need to see, govern, and trust every one running across their business.
Proud to be a founding member of the Blueprint Alliance.
okta.com/newsroom/press-rele…



















