Listen to the audio version of this article (generated by AI).
PPI inflation comes in red hot… a mass cyberattack almost happened… why AI just changed the rules of the game… why legacy cybersecurity stocks might not be up for the challenge… the Prisoner’s Dilemma in effect… what investors should do now…
A disaster.
That’s how legendary investor Louis Navellier described this morning’s Producer Price Index (PPI) report in his Breakthrough Stocks Flash Alert update.
The headline figure jumped 1.4% for the month, crushing the 0.5% forecast, and doubling the revised 0.7% March number.
On a year-over-year basis, the index climbed 6% – the biggest jump since December of 2022.
Let’s go to Louis for what’s behind the numbers:
Here’s the problem: Wholesale goods costs rose 2%. Wholesale service costs rose 1.2%.
That means inflation is increasingly being embedded on the wholesale level and will likely persist.
So, we have Treasury yields going higher this morning. We have the yield curve flattening a bit, and that means all hope for rate cuts are off until we get Treasury yields a little lower.
We need to keep an eye on the services inflation that Louis highlighted – it doesn’t respond quickly to ceasefire agreements or reopened shipping lanes. Once it’s embedded, it takes time to work back out.
It appears that we’re at risk of that happening today.
Despite the hot print, Louis remains bullish – for one simple reason
Earnings.
Back to his update to subscribers:
We should just take a step back and realize that we have 20% earnings growth with the S&P this quarter.
Earnings are forecasted to be good for the remainder of the year. Stocks are a great inflation hedge.
So, despite this news that has rattled the market with inflation, we are in a very good environment.
In all candor, we made too much money too fast, so we are going to have to back and fill here just a bit.
Overall, while hotter inflation may delay rate cuts, Louis doesn’t see it derailing the broader bull market.
He expanded on his outlook during a live event this afternoon, detailing why he believes we’ve entered one of the rarest – and potentially, most lucrative – market windows in decades. This has only happened a handful of times in recent history: 1995, 2001, 2008, 2020 – and now, today.
Louis says it’s the kind of setup that historically produces massive stock winners – and he’s already tracking 53 names showing early signals. Better still, he gave away one of them this afternoon.
If you missed it, you can catch a free replay right here.
Now, let’s shift gears to a story from earlier this week that largely flew under the radar – cybersecurity experts may have just prevented a nightmare scenario…
The massive hack we barely sidestepped
Sometime in the past few months, a group of criminals sat down with an AI model and asked it to break into millions of computers at once.
The AI obliged.
It found a hidden flaw in a piece of software used by businesses worldwide – one that no human researcher had found, and that no traditional security scanner had flagged. Then it wrote the attack code. Clean, methodical, ready to deploy.
This isn’t a hypothetical.
Google’s Threat Intelligence Group (GTIG) confirmed this occurrence on Monday in a report that’s genuinely alarming.
The hacker’s plan was to launch a mass exploitation campaign. Translation – hit as many targets as possible, all at once, before anyone knew what was happening.
Fortunately, Google’s threat intelligence team caught it first, worked with the software maker to patch the hole, and shut the operation down before it launched.
But the next hackers may not make the same mistakes.
What actually happened – in plain English
Without getting too deep in the weeds, the AI-enhanced attack allowed hackers to bypass two-factor authentication. That’s the “enter the code we texted you” step that most of us rely on as our last line of defense. With this exploit, that protection was gone.
What made this different from every previous attack of its kind was how the “skeleton key” that enabled the hack was made. Not by a team of elite hackers working for months, but by an AI model, working for hours.
The GTIG report explained how they knew AI was involved. The attack code had telltale fingerprints – the kind of hyper-organized, over-documented, textbook-perfect formatting that AI models produce when they write code.
From the GTIG report released on Monday:
The script contains an abundance of educational docstrings, including a hallucinated CVSS score, and uses a structured, textbook Pythonic format highly characteristic of LLMs training data (e.g., detailed help menus and the clean _C ANSI color class).
In other words: the AI was so thorough, organized, and eager to explain itself that it gave itself away.
This time.
We’ve entered a new era
For years, this type of sophisticated attack required time and expertise.
It meant months of painstaking work by some of the most skilled people in the world. Nation-states could do it. Elite criminal organizations, occasionally. Everyone else was locked out.
AI has suddenly changed that.
Here’s Ryan Dewhurst, Head of Threat Intelligence at cybersecurity firm watchTowr, in The Hacker News:
AI is already accelerating vulnerability discovery, reducing the effort needed to identify, validate, and weaponize flaws.
This is today’s reality: discovery, weaponization, and exploitation are faster. We’re not heading toward compressed timelines; we’ve been watching the timelines compress for years.
There is no mercy from attackers, and defenders don’t get to opt out.
And it’s not just criminals. The GTIG report documented a sweeping picture of state-sponsored actors already deeply engaged in AI-assisted hacking – China, North Korea, and Russia, all using AI models to accelerate their operations.
We’re talking about attacks that are industrialized, systematic, and happening right now – as you’re reading this.
John Hultquist, chief analyst at GTIG, spoke to The Register about what Monday’s report really represents:
There’s a misconception that the AI vulnerability race is imminent. The reality is that it’s already begun.
For every zero-day we can trace back to AI, there are probably many more out there.
The “Prisoner’s Dilemma” strikes again
There’s a concept we explored in our April 6 Digest that applies directly here – what we called the “Prisoner’s Dilemma” of AI adoption.
The idea is simple…
Every CEO knows that racing to implement AI carries some degree of risk. But if a competitor implements AI and increases efficiency/profitability while that CEO doesn’t, his/her company loses.
So, everyone races – despite the risks. But recognize what this means…
Every time a business wires AI into its operations – connecting its data, its customers, its internal systems to an AI provider – it runs new digital plumbing. But all that plumbing presents a new sea of opportunities for AI-equipped attackers.
The GTIG report provided an example of this…
Earlier this year, a criminal group called TeamPCP snuck malicious code into LiteLLM – a popular piece of software that businesses use to connect their systems to AI providers like Anthropic, Google and OpenAI. Because so many companies had installed LiteLLM, that hidden code quietly stole the digital keys those companies used to access their AI accounts – without anyone noticing.
Think of it like a locksmith who, instead of just making you a key, also made a secret copy for himself. Every customer who came to him for a key got robbed without ever knowing he’d been there.
Nobody at those companies did anything wrong. They were just doing what every company is doing right now – racing to connect to AI before their competitors do.
But the faster these companies race, the more exposed they become to this risk.
This is the Prisoner’s Dilemma in action.
What are the investment implications?
When stories like this break, many investors panic-sell cybersecurity stocks.
In fact, we’ve seen two examples this year – once after Anthropic’s Claude Code Security announcement in February, and again after the Claude Mythos panic in March/April. The market feared that AI would render cybersecurity companies irrelevant.
That’s wrong – AI-powered attacks create more demand for security, not less. But that doesn’t mean every cybersecurity company benefits equally. And that’s where we need to clarify what’s really happening today.
Imagine two home alarm companies…
One built its system from scratch, knowing that thieves would one day use sophisticated technology to case houses. Its sensors don’t just watch for broken windows – they look for unusual patterns of behavior, things that feel wrong even when nothing is technically broken.
The other company took an alarm system from the 1990s – perfectly good at the time – and added a software update to keep up with today’s systems.
When AI-powered burglars arrive, which company’s customers are better protected? And which company gets the new contracts?
Obviously, the companies that were built with AI at their core are structurally better equipped for a world where attackers use AI. The companies that bolted AI onto legacy architectures are scrambling.
So, how should investors respond?
The kneejerk reaction is to reach for a cybersecurity ETF – you get broad exposure with no single-stock risk.
The two most widely held are the First Trust NASDAQ Cybersecurity ETF (CIBR) and the WisdomTree Cybersecurity Fund (WCBR).
While that’s not necessarily wrong, be aware of what this means…
Both ETFs hold a broad basket of cybersecurity names – so, they include plenty of the legacy, bolt-on players we just described. You’d be buying the whole industry at a moment when the industry itself is splitting into winners and losers.
Now, that’s not necessarily a dealbreaker. Wedbush’s channel checks, published earlier this year, found that cybersecurity vendors have set 2026 sales targets up to 30% higher than the typical 12% global spending growth baseline.
So, for now, even bolt-on companies could benefit from a “rising tide” environment as Corporate America spends more on security.
But go in with your eyes open: broad exposure here means, well, broad exposure, which includes the companies least equipped for what’s coming.
For investors who want to avoid this, two of biggest AI-native platforms are CrowdStrike Holdings (CRWD) and SentinelOne (S).
Do your own homework – but start with these two.
We’ve passed the Rubicon
Let’s be candid about what Monday’s GTIG report reveals…
The barrier to launching a sophisticated cyberattack just fell. It won’t go back up.
The companies that understand this – and the investors who see it early – are the ones who will end up on the right side of what’s coming.
We’ll keep you updated as our experts weigh in.
Have a good evening,
Jeff Remsburg
