I’ve been looking into $BRUN because, at these prices, it looked like it could be an attractive AI infrastructure investment. But some of what I found in the filings genuinely shocked me. I started with the cash flow. $BRUN reported $114.2 million of operating cash flow in the first half of 2026. That sounds impressive, until you see that it included a $112.9 million increase in customer deposits. Customer prepayments are useful financing, but they are not the same as recurring cash earnings. Much of that money comes with services still to deliver. Then I followed the money further down the cash-flow statement. BRUN also paid $99.9 million against finance-lease liabilities, outside operating cash flow, alongside $39.7 million of investing outflows. Subtract those from operating cash flow and you get approximately negative $25.4 million. That is my calculation, includes growth investment, and is not a steady-state burn estimate. But it is a very different picture from looking at the operating cash-flow headline alone. At that point, my question changed. I wasn’t asking how quickly $BRUN could announce more contracts. I was asking how much cash would actually remain after paying for the infrastructure behind them. So I looked at the contracts. In the appendix of its Q2 presentation, $BRUN explicitly says its total contract value includes renewal options and letter-of-intent volume that is not yet a performance obligation. That does not establish the composition of its latest contract total. But it means I need a clear reconciliation before treating headline contract value like firm, enforceable backlog. Potential renewals and committed revenue are not interchangeable. The $525.6 million Cohere agreement added another layer. If a specified minimum of infrastructure is not accepted by July 15, 2027, Cohere can terminate and recover its prepayments. Even if that minimum is met, unaccepted capacity can still be cancelled with corresponding refunds. And $BRUN’s listed parent guarantees those unused-prepayment refunds. These are contractual protections, not evidence of a cancellation, but they show how deployment problems could become liquidity problems, rather than simply pushing revenue into another quarter. Then I reached the governance disclosures, and this was where I became considerably less comfortable. $BRUN’s original Q1 filing said its disclosure controls were effective. The amended filing changed that assessment to not effective for the same reporting period, citing material weaknesses. There was no financial-statement restatement, but reversing the assessment of the systems responsible for reliable reporting is something I take seriously. On September 29, the CFO received 2.89 million restricted stock units with a grant-date value of approximately $48.1 million. They vest over time and are not a cash bonus. The company cited his role in the listing and the fact that he held no founder shares. Even with that explanation, an award of that size, following a corrected controls assessment, does not make me more comfortable as a prospective shareholder. None of this proves the business will fail. Customer prepayments can help fund growth, and successful delivery could materially improve the investment case. But I need to see that translate into durable cash generation, not just larger headline contract numbers. I started researching $BRUN because I thought the market might be overlooking an opportunity at this price, but instead I finished questioning whether I was overlooking the risks. $BRUN is a no go for me here.