Senior Republican clears path for next Fed chair

Good morning from the Financial Times. Today is Monday, April 27th, and this is your FT News briefing. Kevin Warsh just cleared a major hurdle to become Fed Chair, and we look at a cybersecurity AI model that could be too effective for its own good. And if this gets into the wrong hands, that could be really dangerous. But first, an update on the alleged Trump shooter. I'm Sonia Hudson, and here's the news you need to start your day.

The man suspected of firing shots during the White House Correspondence dinner is set to be charged in federal court today. U.S. President Donald Trump was attending the Black Tie event on Saturday at the Washington Hilton when multiple shots were heard. City officials said a suspect had run toward the hotel's ballroom entrance, but was quickly apprehended by secret service officers. Authorities say the alleged shooter was targeting Trump and other administration officials, but U.S. Attorney General Todd Blanche said yesterday they're still looking into the motive. Todd Blanche told ABC news that investigators had searched the suspect's home in California. Trump told Fox News that the suspect had written a, quote, manifesto referencing administration officials, and that family members had alerted the police beforehand.

U.S. Republican Senator Thom Tillis said yesterday he'll allow Trump's pick for Federal Reserve Chair to face a confirmation vote. Thom Tillis led a group of senators in blocking Kevin Warsh's nomination. was concerned about a federal criminal investigation into the current chair, Jerome Powell. Thom Tillis and others saw it as part of a larger assault on the central bank's independence. But prosecutors dropped the probe on Friday. Here to tell us what comes next is the FTs U.S. economics editor Claire Jay Jones. Hi, Claire. Hi, Sonia. So, Claire, just to back up for a minute, remind us what this probe was all about. So the Department of Justice said it needed to investigate Fed Chair Jerome Powell's handling of a 2.5 billion renovation of the Fed's headquarters.

So Thom Tillis essentially got what he wanted. He got the Department of Justice to drop this criminal investigation into Jerome Powell. And as a result of that has now said that he is not going to block Kevin Warsh's nomination. Does this mean that Kevin Warsh will get confirmed? I think it almost certainly does. Thom Tillis did have the blocking vote on the Senate Banking Committee. They'll now vote on Wednesday and probably vote in favor of Kevin Warsh moving to a vote. vote on the Senate floor. The sense we're getting is that all of the Republicans who had objections to Walsh, those objections were really centering on the probe. And now that the probe looks as though it's gone away, they're likely to support Walsh. There's 53 Republicans in the Senate at the moment. So it looks like war shall pass the crucial hurdle of 51 quite easily. Claire, something I'm wondering about is, is there anything stopping the Trump administration? from reopening the investigation into Jerome Powell after they get what they wanted, which is

Kevin Warsh getting confirmed. So the U.S. Attorney for the District of Columbia, Janine Piero's remarks on Friday did raise a little bit of concern because she did mention in a post on social media site X that she will not hesitate to restart a criminal investigation should the facts warrant doing so. So the sense we've got from remarks that Thom Tillis made yesterday was that he'd had assurances from the DOJ that the only way the probe would be reopened was if the Fed's internal Inspector General recommended that it do so. So he seems to have been assured enough by conversations with the DOJ to think that they will not reopen the probe, despite the being some gray area created by Piero's remarks on Friday.

Claire, this has been quite a saga, and, you know, we haven't even touched on some of the other ways that Trump has potentially attacked the independence of the Fed. What kind of environment is Kevin Warsh stepping into, you know, if and when he does take the helm at the Fed? Well, I think, as you say, Walsh has taken the helm at a time when the Fed is under the sort of political. pressure, which I think it's never really experienced. We've seen the attempt to take legal action against Jerome Powell. We've also seen the US president attempt to fire. Fed Governor Lisa Cook over allegations of mortgage fraud, which he also denies. And Cook remains at the Fed. However, you could argue that the resistance of Jerome Powell and Cook and the sense in which the Fed has had some success in the courts does provide cover for Walsh and that, in a sense, he is going to have a few legal

protections that we now know about that maybe we weren't quite aware of at this time last year. Claire Jay Jones is the FTs U.S. economics editor. Thanks, Claire. Thanks, Sonia.

Speaking of central banks, several major ones are meeting this week. They include the Fed, the European Central Bank, and the Bank of England. They're all expected to hold off on raising interest rates. You see, the energy price shock from the U.S.-Iran war is making inflation harder to predict. That's partially driven by social media posts from President Donald Trump. Traders have struggled with the volatility caused by his frequent posts and the responses from the Iranian regime. Monetary policymakers are now relying less on single forecast. Instead, they're focusing on scenarios that take into account a range of possible outcomes in the Middle East conflict.

Artificial intelligence is massively boosting cybercrime. It gives amateur hackers cheap tools to write harmful software, and it also allows a criminal to automate and scale their operations. Enter Anthropics Claude Mythos model. Some people are worried it's so good, it could actually backfire if it gets in the wrong hands. And last week, the AI lab started investigating whether it was accessed by unauthorized users. The FT's Christina Criddle is here to tell us more. Hi, Christina. Hi.

So what was Anthropics setting out to build with this cybersecurity AI model called Mythos? So Anthropic didn't actually set out to build a cybersecurity specific model. it was just building its next big general purpose AI model in its Claude series, and it then discovered in testing that it was very capable at cybersecurity. And in particular, it was good at detecting bugs or flaws in software, but also this new capability of being able to generate exploits, ways to work around those bugs or flaws and exploit them. And that's why it decided to only release it to a select amount of partners.

Yeah, and those partners include Amazon, Microsoft, and Cisco. Tell me a little bit more about concerns that have been raised about this model. So AI models have been able to do this for some time now, but I think Mythos really was a lightning rod. And the idea that AI can scan for these vulnerabilities and bugs should be a positive one, right? So you can spot where all the flaws are and then go and fix them. But the issue is a volume problem where the AI means that you're potentially detecting more bugs than you can solve for at a particular time. That means that they're left vulnerable to attack by hackers and potentially foreign adversaries. So earlier I mentioned this potential unauthorized breach of mythos.

What concerns is that raising? Yeah, I mean, I would just say there is some skepticism about how elevated clawed mythos is with cybersecurity. compared to what we've seen before. But I definitely think this is a moment where wider sectors and industry and governments are recognizing this as a problem. And then to add on the pressure here of the threat, Anthropic is investigating this report

which claimed that a group of people had unauthorized access to Claude Mythos. Apparently, they were able to access the model through a third-party vendor environment as one member of the group apparently was a contractor for Anthropic. And in the run-up to Mythos, there were already two security issues. So there is this criticism that maybe Anthropic doesn't have its own security in order. And what does that then mean if unauthorized people are able to access this model, which they claim is so powerful. And if this

gets into the wrong hands, that could be really dangerous. Well, what has Anthropic said about these issues? Anthropic has said that they're investigating this report, which claims unauthorized access, They say it's through one of this third-party vendor environments and not through its API, which is how most developers would access this model. And in terms of the previous two security issues in the run-up to Mythos, they've said this was human error, this was not caused by AI. But that still raises quite a big question about how secure their infrastructure is. Christina Criddle is a tech correspondent for the FT.

Thanks, Christina. Thank you. You can read more on all these stories for free when you click the links in our show notes. This has been your daily FT News briefing. Check back tomorrow for the latest business news.

Capital raising is being redefined in real time, from macro disruptions and shifting investor expectations to AI-driven demand and the rise of private credit. When headlines can tilt markets overnight, how can companies stay on course? Strategic Alternatives, the RBC Capital Markets Podcast, explores how corporates and investors are navigating new pathways to raise capital, create value, and drive growth. When conditions change, alternatives matter. Listen to strategic alternatives, available wherever you get your podcast. The latest episode of the next five podcast is all about logic and trust in AI. I speak with Scott Will to move at AWS.

With the Gen. AI, there's always an error rate. Automate reasoning, there's never an error rate. Mary Martin at BCG. It is culture, not code. Organizational inertia is an incredibly powerful force. And Vavarin Chickcara at ADP. What we should be teaching the next generation is this form of discipline judgment. You can listen to the full episode of the next five, wherever you get your podcasts.