All Posts

The consulting industry is experiencing one of the most significant internal shifts in its modern history, a shift driven by a rapid acceleration in artificial intelligence spending combined with organizational strategies that are struggling to justify it. Over several years, the largest firms invested extraordinary amounts in new platforms (AI), training programs (For AI), and automation initiatives (Again, AI), creating an internal expectation that employee behavior must no

The latest reporting from Reuters on the surge of Iranian‑aligned cyber operations following U.S.–Israeli strikes makes one thing clear: the digital spillover from geopolitical conflict is no longer a theoretical risk, it is an operational inevitability. When state pressure rises in the Middle East, the first shockwaves are felt not in government networks but across the commercial sector, where visibility is low, governance is inconsistent, and attackers know disruption is ea

I was recently asked to attack vector (how or where I would compromise) an AI system to maximize damage, so I thought to share my thoughts. AI systems are only as trustworthy as the data that trains, tunes, and feeds them. Yet the weakest link in that data pipeline isn’t the model, the cloud, or the enterprise perimeter, it’s the down‑supply chain, where third or fourth party vendors, contractors, and open‑source dependencies quietly shape the inputs that AI systems rely on.

Central Bank Digital Currencies (CBDC’s) are often misunderstood because they get lumped into the same category as crypto, stablecoins, and digital assets. But a CBDC is something fundamentally different. It is not a token, not a market product, and not a speculative instrument. A CBDC is simply the national currency itself, issued in digital form by the central bank. What makes CBDCs transformative is not that they are “digital.” Money has been digital for decades. The shif

The Oracle news is not an isolated headline; it is the clearest signal yet that the AI bill is coming due and the timing could not be worse for firms already stretched thin. When a company of Oracle’s scale begins preparing layoffs that could reach 30,000 people while managing the weight of a $300 billion OpenAI commitment, it exposes the structural truth most executives still refuse to say out loud: AI is a front‑loaded cost with a back‑loaded payoff, and the gap between tho

Companies cutting staff while accelerating build velocity under the banner of AI are making a catastrophic category error: they’re assuming AI will self‑learn its way into stability. It won’t. AI doesn’t self‑govern, self‑audit, or self‑maintain; it simply amplifies whatever system it’s dropped into. And when you remove humans, you remove the only force that counteracts entropy. The result is predictable: the system begins to expand faster than anyone can understand, secure,

The industry is collapsing two distinct eras of AI into one, and it’s distorting our understanding of where we actually are in the evolutionary arc. Everyone talks about “AI” as if it’s a single monolithic capability, but generative AI and agentic AI represent fundamentally different phases with different risks, economics, and governance requirements. Generative AI is the era we are living in today. It is the ability to produce: text, images, code, video, audio. These systems

The AI sector is being celebrated as the next industrial revolution, but a more uncomfortable truth is emerging: much of today’s “growth” is not organic demand, it’s circular financing. Companies are investing in their own customers, who then spend that money right back with them, inflating valuations and creating the illusion of unstoppable momentum. Nvidia is the clearest example. The company has poured billions into OpenAI and Anthropic, firms that immediately turn around

The AI ecosystem continues to accelerate in ways that feel less like market evolution and more like engineered inevitability. NVIDIA’s newly announced one‑gigawatt partnership with Thinking Machines Lab is the clearest signal yet that the company is no longer just selling computing, it is architecting the entire upstream and downstream flow of capital, infrastructure, and influence. In my last article, I broke down the circular investment loop where NVIDIA funds AI labs, thos

I’ve been posting high‑level warnings about the current AI narrative, but it’s time to go deeper for the technical audience. The hype cycle around AI, especially agentic AI, has accelerated to a point where the engineering reality is being overshadowed by marketing language. I’m a strong supporter of AI and its long‑term potential, but enthusiasm without skepticism is how institutions expose themselves to operational, financial, and regulatory risk. A large language model (LL

The industry keeps promising autonomous agents, but the real constraints aren’t model size or clever prompting. They’re power, cost, and the physical limits of the infrastructure required to run them. Agentic AI demands persistent memory, long‑horizon planning, and continuous tool use, all of which require massive, uninterrupted compute. The global grid wasn’t built for this. Data centers already consume more electricity than some nations, and the next generation of agentic s

Most organizations are preparing for AI threats at the perimeter, focusing on prompt injection, data leakage, and hallucinated outputs. The real danger sits deeper, inside the operational workflows where AI is quietly granted trust it never earned. Here is a red team scenario that exposes that blind spot, mostly takes patience and even if not deployed 100% correctly will still do some damage. Scenario: The Infiltration Layer Your enterprise deploys an internal AI system (mo

The attack on Stryker is a textbook example of how a modern, state‑aligned adversary turns a single architectural weakness into a global operational collapse. While Stryker has not disclosed the precise entry vector, the observable blast radius tells the story: the attackers gained privileged access to Stryker’s Microsoft ecosystem, leveraged that foothold to compromise identity infrastructure, and then used the company’s own device‑management tools to deploy destructive payl

The accounting profession is entering the same phase the tech sector hit two years ago: the quiet part is being said out loud. Firms aren’t just “rightsizing.” They’re pre‑paying for an AI future they haven’t actually built yet. PwC has cut nearly 3,300 roles across 2024–2025, concentrated in audit and tax, the very functions now being re‑architected around automation and low‑attrition surpluses. EY, Deloitte, and KPMG have all followed with their own reductions, citing effic

For years, penetration testing was treated as the gold standard of cybersecurity assurance. Today, it has quietly become one of the most over‑purchased and least‑understood services in the entire industry. And the uncomfortable truth is this: most organizations are paying for something that barely scratches the surface and speaks more to an antiquated notion of the perception of security more than any real semblance of it. Let’s acknowledge the reality first. Some compliance

A zero‑day is the most dangerous category of vulnerability in cybersecurity: a flaw the vendor doesn’t know exists, has no patch for, and cannot defend against. “Zero‑day” means defenders have had zero days to prepare. The attacker finds the weakness first, weaponizes it, and bypasses every control built on historical patterns or known signatures. The recent Cisco Secure Firewall Management Center (FMC) zero‑day is the clearest example of how modern cyber risk has shifted fro

Instagram’s decision to eliminate end‑to‑end encrypted messaging by May 8, 2026 is not a routine product update. It is a structural reversal of one of the last remaining privacy protections on a platform already engineered to harvest and monetize human behavior at scale. Meta’s claim that “very few people opted in” is not a real explanation. It is an admission that encryption was inconvenient to the business model. End‑to‑end encryption is the only mechanism that prevents Met

Every institution is quietly confronting the same truth: our digital identity infrastructure is collapsing, and no one wants to say it out loud. We built an entire financial system on the assumption that identity could be verified, authenticated, and trusted. That assumption no longer holds. The controls we relied on, KYC, document verification, behavioral analytics, device intelligence, were designed for a world where fraud was linear and human. Today’s fraud is synthetic, a

Every week, another headline announces a massive AI deal like Nvidia selling a million chips to Amazon by 2027. But if you strip away the hype, what you see isn’t organic demand. It’s a closed‑loop financial choreography designed to keep valuations inflated while the real economics of AI remain unproven. Hyperscalers buy billions in Nvidia hardware not because customers need it, but because Wall Street demands the appearance of acceleration. Nvidia books record revenue. Amazo

The audit profession is facing an existential problem it refuses to name: AI has outpaced the very assurance model the industry is built on. For decades, auditors have been trained to evaluate evidence, not the systems that generate it. That distinction mattered when evidence was physical, scarce, and difficult to fabricate. It does not matter now. AI can generate invoices, contracts, bank statements, payroll records, vendor files, and entire financial histories that are inte