Rob Knott

We recently announced an upcoming report on the government finance function strategy to 2030. Find out more ⤵️ nao.org.uk/work-in-progre…

How Universal Credit went wrong - in one graph...

I see some weird things but this takes the biscuit. A vulnerability in the Companies House website, that let anyone view the private dashboard of any one of the five million registered companies, see directors' personal details. And modify them.

Tory peer quits after Lords probe into PPE deals…. @HCSAprocurement bbc.co.uk/news/articles/…

And a reminder of the latest update from @NAOorguk

What does the new £37 billion Hospital 2.0 Alliance framework mean for the New Hospital Programme and NHS supply chains? @HCSAprocurement supplychaindigital.com/news/hospital-…

We've published our latest overview of the Department of Health & Social Care. Our 2024-25 report looks at how @DHSCgovuk spent £219.2 billion, its plans to improve NHS productivity, wider changes and key risks facing the department. Our overview ⬇️ nao.org.uk/overviews/depa…

Sopra Steria sues UK government over £958M Capita outsourcing award dlvr.it/TRBgdX

Anthropic just announced it will take the Trump administration to court over the supply chain risk designation. And in the same breath, Axios revealed the detail that changes everything about this story. While Anthropic was being blacklisted for refusing to allow mass surveillance, the Pentagon’s own “compromise deal” that Under Secretary Emil Michael was offering on the phone at the exact moment Hegseth posted the designation on X would have required Anthropic to allow the collection and analysis of Americans’ geolocation data, web browsing history, and personal financial information purchased from data brokers. Read that again. The Pentagon spent two weeks saying it has no interest in mass surveillance of Americans. Then the deal they actually put on the table asked for access to your location, your browsing history, and your financial records. They told us Anthropic was lying. The contract language told us Anthropic was right. Now here is where this becomes an existential question for a $380 billion company. The supply chain risk designation means every company that does business with the Pentagon must certify they do not use Claude. Eight of the ten largest companies in America use Claude. Defense contractors, cloud providers, consulting firms, banks. The blast radius is not the $200 million Pentagon contract. It is the enterprise ecosystem that generates $14 billion in annual revenue. Anthropic’s legal argument is specific: under 10 USC 3252, the designation can only restrict use of Claude on Pentagon contract work. Your commercial API access, your claude.ai subscription, your enterprise license are, in Anthropic’s reading, completely unaffected. But here is the problem. That is a legal argument. It will take years to resolve in court. And in the meantime, every general counsel at every Fortune 500 company with any Pentagon exposure is going to ask one question: is using Claude worth the risk? The IPO, which was expected this year at a $380 billion valuation backed by $30 billion in fresh capital, is functionally frozen. No underwriter will price an offering while a company carries the same designation as Huawei. And here is the final detail nobody has processed yet. Hours after blacklisting Anthropic, the Pentagon accepted OpenAI’s proposed safety framework, which contains the identical red lines: no mass surveillance, no autonomous lethal weapons. They destroyed one company for a position they then accepted from its competitor. Full analysis on Substack. open.substack.com/pub/shanakaans…

Perplexity just became the the first Al company to truly go head-to-head with the Bloomberg Terminal... Using Perplexity Computer (with no local setup or single LLM limitation), it was able to build me a terminal with real-time data to analyze $NVDA using Perplexity Finance:

Meet Simon Goldberg. He tells people the US Govt will pay your UK gas, electricity, credit card bills - all your spending. Yes, it's fraud. But the IRS actually sends the cheques. The insane part? He started a *private prosecution* of the YouTuber who revealed it. 🧵

A blog post just wiped $30 billion off IBM in a single afternoon. Not a product launch. Not an earnings miss. Not a competitor undercutting on price. A five-minute blog post explaining that Claude can read COBOL. IBM dropped 13%. Worst single-day loss since October 2000. Twenty-five years of stock resilience ended by one AI company publishing a capability update. Here’s what happened: 95% of ATM transactions in America run on COBOL. Hundreds of billions of lines power banking, airlines, and government systems. The developers who built them retired decades ago. The knowledge left with them. Finding engineers who can even read COBOL gets harder every quarter. IBM’s moat was never the technology. It was the fact that nobody else could understand it. Entire consulting empires existed because the code was too old, too tangled, and too critical to touch. Companies paid IBM billions because the alternative was catastrophic system failure. Then Anthropic published a blog post saying Claude Code can map dependencies across thousands of lines of COBOL, document workflows, identify migration risks, and translate legacy logic into modern languages. Modernization in quarters instead of years. The market heard: the priesthood just lost its monopoly on the sacred language. And this isn’t the first time. Last week Anthropic announced Claude Code Security for vulnerability scanning. CrowdStrike dropped. Okta dropped. Cloudflare dropped. One company is serially destroying legacy moats with blog posts. Now here’s where it gets surreal. This same company, on the same day, also published evidence that three Chinese AI labs ran 24,000 fake accounts and 16 million exchanges to steal Claude’s capabilities. DeepSeek used it to build censorship tools. MiniMax pivoted within 24 hours when a new model dropped, redirecting half its traffic to steal the latest version. And yesterday, the Pentagon summoned this same company’s CEO for what officials called a “sh*t-or-get-off-the-pot meeting,” threatening to blacklist them like Huawei for refusing to let the military use Claude without safety restrictions. Three stories. One company. Twenty-four hours. The company destroying legacy moats faster than the market can reprice them is simultaneously being threatened by its own government and looted by foreign competitors. Anthropic is valued at $380 billion. Its CEO says a 12-month delay in AI would make him bankrupt. The Pentagon wants to designate it a supply chain risk. Chinese labs are running industrial espionage against it. And it just proved it can vaporize $30 billion in market cap with a Monday morning blog post. Whatever you think about AI disruption, IBM’s stock just settled the argument. Full institutional analysis on my Substack. open.substack.com/pub/shanakaans…

How Palantir captured the Ministry of Defence…. ft.com/content/520792…

I voted yes on a $100 billion investment. It took forty minutes. Seventeen of those minutes were lunch. I did not read the memo. The memo was 94 pages. Page 1 said "Artificial General Intelligence." Page 94 said "$830 billion valuation." Everything in between was a graph going up and to the right. The graphs measured "Total Addressable Intelligence." That is not a real metric. But it had a three-letter abbreviation. TAI. Three-letter abbreviations are how you know something is serious. The revenue was $5 billion. The loss was $14 billion. That's $2.80 lost for every dollar earned. I called it "investment-phase economics." The other partners liked that phrase. One of them wrote it down. One of them underlined it twice. Someone showed the path to profitability. A line starting at $5 billion and ending at $280 billion. By 2030. That's a 56x increase in four years. The line was very steep. Steep is good in a pitch deck. Steep means ambition. Flat means you have a business. Nobody wants a business at this valuation. Thirty-seven investors signed on. None of them asked how $5 billion becomes $280 billion. You don't ask that at $830 billion. You nod. I nodded. The company used to be a nonprofit. Its founding mission was "ensuring AI benefits all of humanity." We converted it to for-profit. I asked what happened to all of humanity. Someone said they were "grandfathered in as stakeholders." I asked what that meant. He said "it means they're in the deck." Slide 11. Small font. The infrastructure plan calls for $600 billion in data centers. I asked where $600 billion would come from. Someone said "future rounds." I asked what would fund the future rounds. Someone said "the infrastructure." I did not ask a third question. The third question is where the math breaks. You never ask the third question. The CEO told Congress that AI safety requires responsible governance and significant investment. He told investors that AI dominance requires speed and significant investment. Same week. Same billions. Different adjective. Congress got "responsible." Investors got "dominant." Both rooms nodded. Nodding is the primary output of an AI investment meeting. We are very good at it. A junior analyst at my fund ran the numbers independently. She calculated that at the current loss rate, we'd need to raise $200 billion more before reaching profitability. She put this in a memo titled "Structural Concerns." I promoted her to Head of Strategic Foresight. She forecasts now. She doesn't calculate. There's a difference. Calculating uses numbers that exist. Forecasting uses numbers that should exist. We prefer the second kind. I tried the product once. ChatGPT. I asked it to summarize the investment memo I had not read. It got three of the numbers wrong. One of them was the valuation. The valuation it hallucinated was lower than the real one. The real one was less believable. I voted yes anyway. A reporter asked me if an $830 billion valuation for a company losing $14 billion a year made sense. I said "we're not investing in a company, we're investing in a paradigm." She asked what that meant. I said "exactly." She printed it. It sounded profound in the article. It wasn't. But at this valuation, everything sounds profound. The LPs in my fund asked about the OpenAI position. I showed them a chart titled "AI Market Penetration: 2026-2035." I generated the chart with ChatGPT. It hallucinated two of the data points. I left them in. They improved the trajectory. The LPs nodded. Nobody checked the data points. Nobody checks the data points. The data points are not the point. The point is the trajectory. The trajectory is the story. The story is the raise. And the raise is $100 billion. Microsoft invested $13 billion before us. They integrated the product into everything. Adoption is early. "Early" and "low" mean the same thing. But "early" sounds intentional. I'm told it will accelerate. Acceleration is always in the next quarter. The next quarter is always soon. Soon is not a timeline. It's a feeling. Feelings close rounds. We are now in the largest private funding round in the history of technology. For a company that loses $2.80 on every dollar it makes. With a plan to grow revenue 56x in four years. While spending $600 billion on infrastructure that doesn't exist yet. After converting from a nonprofit whose mission was to benefit all of humanity. The metrics are extraordinary. None of them are financial. All of them are directional. Directional means they point in a direction. The direction is up and to the right. Up and to the right is all you need. I'll make Managing Partner this year. My thesis was simple: "AI is inevitable." Inevitable means you can't be wrong. If the company succeeds, I saw it coming. If it fails, the market wasn't ready. Either way, I saw it coming. That's the beauty of inevitability. It's the only investment thesis that can't be disproven. I still don't know what a token is. But I know what the round raised. And I know what the next round will raise. The second number is always larger. That's the only math that matters.

The New York Times made news the loss leader for a $2 billion digital revenue machine, and this chart is the receipt. News-only subscribers dropped 65% since June 2022. Bundle subscribers grew 227%. That looks like a news collapse. But the NYT deliberately killed its standalone news product. They stopped marketing it. They made it nearly impossible to buy a news-only subscription on their website. They priced the full bundle (News + Games + Cooking + Athletic + Wirecutter) at $2/month introductory, cheaper than a standalone Games subscription. News-only ARPU is $13.33. Bundle ARPU is $12.92. Single non-news product ARPU is $3.36. Those 4.3 million single-product subscribers paying $3.36/month? They’re not the business. They’re the funnel. The NYT CEO said it explicitly on the earnings call: single products are “funnels to get people to subscribe” to the bundle. Games now accounts for over 50% of time spent inside the NYT app. Wordle, Connections, and the Mini pull 10+ million weekly players who never intended to read a news article. But half of all NYT subscribers now pay for the bundle, and bundle subscribers retain longer, engage more, and accept price increases. The bundle just went from $25 to $30/month. The result: digital revenue crossed $2 billion for the first time in 2025. Free cash flow hit $550 million. Adjusted operating margins reached 24% in Q4. Berkshire Hathaway just took a billion-dollar position. While the Washington Post cut 300 journalists last week, the Times added 1.4 million subscribers. This chart shows a news company that built an attention ecosystem where Wordle gets you in the door, Cooking keeps you at breakfast, The Athletic owns your commute, and by the time you think about canceling, you’d lose four products instead of one. The NYT figured out that the way to fund journalism in 2026 is to make sure you can’t quit the crossword.

@Gabriel_Pogrund They broke into a broken safe. Sounds about right.

EXCL: The Cabinet Office’s propriety and ethics team broke into a safe and destroyed the department’s copy of a historic bullying investigation into Antonia Romeo, the new head of the civil service. thetimes.com/uk/politics/ar…

The art of the deal... After negotitations with big-name drug manufacturers, TrumpRx claimed 'the world's lowest prices on prescription drugs' But - compared to self-pay market in the UK, TrumpRx is *more expensive* on 4 out of its 6 headline drugs lawrencedunhill.substack.com/p/how-trumprx-…

@HSJEditor And interesting how Palantir has weaved itself into the MOD supply chain alongside its direct contracts….

Looks smarter than the FDP

Firm assessing Covid vaccine harm replaced after costs spiral to £48 million…. @HCSAprocurement bbc.co.uk/news/articles/…