The Register Kettle

Google I/O has ostensibly been an AI show for a few years running, but this year's announcements have taken the cake, which Google seems all to happy to let its users eat as it reshapes the web. On this week's episode of The Kettle, host Brandon Vigliarolo is joined by El Reg senior reporter Tom Claburn and open source reporter Liam Proven to discuss how Google's bevy of AI announcements, and declaration that we're entering the era of AI search, might not play well with customers. From an enlarged AI mode, to AI ads stuffed into AI answers, and pushing AI devs onto closed-source tools after shuttering open-source ones, Google is leaning hard into its version of the future of the internet no matter what users might think, and we wonder whether that might finally crack Google's stranglehold on the web. 

Hopefully you haven't had reason to notice yet, but there's a rising problem with AI services on Google Cloud, AWS, and other platforms sticking their customers with bills in the tens of thousands of dollars. This week's episode of the Kettle focuses on two such stories that The Register published this week, one concerning Google and another involving AWS. In both cases cloud customers using AI incurred massive bills without any prior notification from their provider and not a lot of help to resolve the matter with any sense of urgency. Tune in to this week's episode to hear host Brandon Vigliarolo chat with O'Ryan Johnson and Richard Speed chat about their stories, what's causing these massive bills, and how you can avoid a similar situation at your own organization.

This week on The Kettle, host Brandon Vigliarolo is joined by systems editor Tobias Mann and senior reporter Tom Claburn to discuss the current state of locally-installed coding assistant LLMs. After some experimentation, both Claburn and Mann have concluded they're actually getting pretty good.What does that mean for the computing power crunch? Hopefully it means cash-strapped developers will be able to get off the train of increased prices from AI vendors and actually run a useful local code assistant - just don't expect to give up on the big guys entirely - you might still need them to okay that smaller AI's work.

Windows is a mess, Copilot's been a disaster, updates have been failing ... hey, Microsoft - everything okay?There's been a lot going wrong for the Windows maker of late, and leadership is promising improvements. Join host and El Reg Brandon Vigliarolo, US editor Avram Piltch, and Microsoft reporter Richard Speed to pick apart the situation and read the entrails to see what Microsoft's future holds.

This week on the Kettle we split the conversation between two topics: The biggest news out of Google Cloud Next, and the latest hype-bursting revelations about Anthropic's Mythos AI, and the fact that it's already been accessed inappropriately.Different topics, yes, but both point to the same thing: AI is reshaping everything from Google's hardware designs to the security world, and we'd better get ready: The tech world isn't slowing down for anyone.

Prompt injection attacks are, in a lot of ways, like phishing a human: By convincing them of your harmless intent, they do something that's not in their best nature just like the gullible person handing over corporate details. Thinking about it that way, our current AI paradigm will likely never escape prompt injection attacks - you can't patch out the ability to con someone with language, human or not, it seems. Join host Brandon Vigliarolo, El Reg cybersecurity editor Jessica Lyons, an senior reporter Tom Claburn to talk about just what prompt injections mean for the future of AI and cyberseurity.

Anthropic stunned the infosec world this week with the announcement of a new zero-day discovering and exploiting AI called Mythos, which it claims is too dangerous to release to anyone but the biggest companies in the tech world.Join host Brandon Vigliarolo, APAC editor Simon Sharwood, and senior reporter Tom Claburn to discuss the Mythos news, what it means, and why it's probably appropriate to be a bit skeptical of all the company's claims.

A human mistake at Anthropic last week led to the entirety of Claude Code's source code being leaked to the web, pubished on GitHub, forked, dissected, picked apart, and otherwise exposed to the world to see - perfect fodder for a Kettle of vultures looking for some news to pick apart.Join cybersecurity editor Jessica Lyons, senior reporter Thomas Claburn, and host Brandon Vigliarolo as they pick apart the leak of Claude Code's source, its security implications, what's been found in the code, and more.

Did you know that if you tell your AI it's an expert coder it'll actually write worse code? Join Brandon, systems editor Tobias Mann, and senior reporter Thomas Claburn to discuss the state of - and limitations with - AI software development, and why we're not convinced it's going to replace a human dev anytime soon.

We regret to inform you that RSAC 2026 is probably going to be all about agentic AI, and with good reason: It's changing both the defensive and offensive security paradigms. Along with agentic AI, The Register security editor Jessica Lyons tells host Brandon Vigliarolo that other hot topics at the industry's big event this year are likely to include the war in Iran, North Korean IT employment scammers, and maybe a bit of politics too: The US government won't be represented at all at this year's event at all.

Join host Brandon Vigliarolo, El Reg's systems editor, Tobias Mann, and US editor Avram Piltch to discuss what we think might be on deck for this year's Nvidia GTC, starting Monday, March 16.

Join host Brandon Vigliarolo, The Register cybersecurity editor Jessica Lyons, and systems editor Tobias Mann to discuss how the US/Iran conflict seems like a new era in the primacy of tech's role in war, how an international conflict is shaping nascent tech hotspots, and more.

The Kettle returns with a new format (audio only) and new host (ugh, he's American?), but the same old banter, analytics, and insight it had before the break. Join us for our first episode, coming soon.