--- title: "Mark Cuban says AI will split companies into winners and bankrupt losers. He's mostly wrong." description: "On 5 May 2026, Mark Cuban told Benzinga that AI will split companies into winners and those that go out of business. The clip travels well. The framing is wrong for almost every UK small and mid-sized business. The real split is winners and the quietly stagnating, and that's a slower, more dangerous problem than bankruptcy because nobody calls a board meeting about it." canonical: https://richardbatt.com/blog/mark-cuban-ai-winners-losers-contrarian date: 2026-05-05 author: Richard Batt tags: [AI Strategy, Contrarian, SMB AI, UK Business] type: blog_post --- # Mark Cuban says AI will split companies into winners and bankrupt losers. He's mostly wrong. _On 5 May 2026, Mark Cuban told Benzinga that AI will split companies into winners and those that go out of business. The clip travels well. The framing is wrong for almost every UK small and mid-sized business. The real split is winners and the quietly stagnating, and that's a slower, more dangerous problem than bankruptcy because nobody calls a board meeting about it._ **Richard Batt** — AI implementation specialist. 120+ projects across 15+ industries, serving SMBs (5-200 employees) worldwide from Middlesbrough, UK (working globally). Contact: richard@richardbatt.com · https://richardbatt.com On 5 May 2026, Mark Cuban told Benzinga that AI is about to split the corporate world into two groups: winners and "those that go out of business." It's a tidy line. It clipped well on LinkedIn. By Tuesday afternoon, three different agency owners had emailed me about it. Two of them sounded a bit panicked. If you run a small or mid-sized business (SMB) in the UK and you watched that clip, here is the practitioner answer. **The short version** - Cuban's binary makes a great soundbite. It misdescribes what actually happens to most SMBs that ignore AI. - Bankruptcy is the loud failure. Quiet stagnation is the common one, and it costs more in total. - A 30-person services firm I worked with last month wasn't going bust. They were just losing 14% gross margin a year to faster competitors and didn't notice until the renewal call. - The honest split is winners versus plateauers, not winners versus losers. - The fix is one workflow this quarter, not a 24-month transformation programme. After 120+ AI projects across 15+ industries, mostly UK SMBs in the 5 to 200 employee range, I have not yet seen a single business go bankrupt purely because it skipped AI. But I have seen plenty of them lose ground quietly. That distinction matters because the prescription is different. ## What Cuban actually said Benzinga's 5 May piece quotes Cuban saying AI adoption will produce companies that thrive and companies that go out of business. He is talking about a five to ten year horizon and he is mostly speaking to listed mid-caps and large enterprises. And read in that frame, the line is not crazy. Public companies have analysts, activist investors, and quarterly earnings calls. If their margins compress for four quarters, the share price moves and the board moves with it. Bankruptcy can follow. The problem is that the clip travels without the frame. By the time it reaches a 40-person Sheffield engineering firm, the binary has been stripped of context. The owner reads "winners or out of business" and decides he needs a rescue plan when what he actually needs is a more boring conversation about throughput. ## Why the binary misleads SMBs Three reasons the clip is the wrong shape for the businesses I work with. ### 1. Most SMBs don't go bust. They plateau. Public-company death is fast. SMB death is slow, and it usually happens for reasons that have nothing to do with AI: a key client leaves, the founder gets ill, a recession hits the supply chain, the lease comes up for review. AI sits underneath all of those things, but it is rarely the visible cause. What I see far more often is the plateau. The firm keeps the same revenue for three years while a faster competitor compounds at 8 to 12% a year on the same customer base. After three years of compounding, that competitor is half a size larger and bidding on contracts the original firm cannot match on price. Nobody filed for administration. The original firm just got quietly outgrown. A 30-person professional services firm I worked with last month is the textbook version. They had not lost a client in two years. Their revenue chart looked flat and stable. When we measured against three direct competitors, those three had each grown 11 to 18% in the same window using AI-assisted proposal writing, faster onboarding, and same-week reporting. The firm I was advising had lost roughly 14% gross margin to those competitors on three contract renewals because their proposals took two weeks longer to turn around. No bankruptcy. Just a slower, less obvious bleed. If you tell that owner he is on track to "go out of business," he tunes out, because he can see his bank balance and he is fine. If you tell him three of his competitors are quietly compounding past him, he leans in. The framing matters. ### 2. The winners aren't the ones with the biggest AI budgets Cuban's binary implies a spending race. The companies on the wrong side of the line, in his framing, are the ones that didn't invest enough or fast enough. That is not what the data says. The Gallup 2026 State of the Global Workplace report, published in April, surveyed AI-adopting firms and found 65% of US workers report personal productivity gains. Only 12% strongly agree their company has actually changed how it operates. Most of the spending money went into individual productivity that never made it onto the P&L. The companies that closed the gap were not the highest spenders. They were the ones whose managers redesigned one workflow and reinvested the saved capacity. Gallup put a number on that effect: employees with a manager who actively supports AI are 98.7 times more likely to say AI has transformed how work gets done. The win there had nothing to do with budget. It came from managerial conviction. Cuban's binary suggests the winners outspent the losers. The data says the winners outshipped them on a much smaller budget, because they redesigned the work instead of just buying tools. ### 3. The cost of inaction is real but slow I am not arguing AI doesn't matter. It does. I am arguing the cost shows up gradually and on metrics most owners aren't watching closely. Five things I see in firms that haven't started: 1. Proposal turnaround drifts from 4 days to 7 days while competitors hold at 3. 2. Customer support tickets accumulate a 12 to 18 hour first-response lag. 3. Onboarding new clients takes 50% longer than it did three years ago, and nobody can quite say why. 4. Junior staff burnout rises because the senior team can't afford to delegate the boring half of every project. 5. Margins on smaller contracts compress because the firm can't profitably take work under £25,000. None of those is a bankruptcy event. All of them are real and costly. A firm losing 14 to 22% of margin to slower turnaround over three years is not a winner under Cuban's frame, but it is also not "going out of business." It is just having a worse year than it should have, every year, until it eventually sells to a competitor at a multiple it would have hated three years ago. ## The honest version of the split If I had to write the Cuban line for SMBs, it would go like this. Most UK SMBs will not go bankrupt because of AI in the next five years. A meaningful minority will compound past their competitors by redesigning one workflow per year and reinvesting the saved time. The majority will plateau, lose 1 to 3% of margin annually to faster firms, and notice somewhere between year three and year five. Some of those plateauers will eventually sell, retire, or restructure under pressure. None of that is dramatic. All of it is real. That framing leads to a different conversation than Cuban's. Instead of "we need to transform or die," the question becomes "which one workflow do we redesign this quarter, and who owns the redesign?" The answer is almost always the same. Pick the workflow with the highest manual hours per week. Pick the manager who is genuinely curious. Give them six weeks to redesign the workflow, then measure the throughput change against the baseline. That's the entire prescription. It doesn't make a clip on LinkedIn. It does close the gap. ## What to do this week If the Cuban quote landed and you want to do something useful with the energy, three actions: 1. Pick one workflow your team does manually for more than 4 hours a week. Proposal writing, invoice extraction, customer support tier one, onboarding emails are the usual candidates. Don't pick five. Pick one. 2. Talk to the manager who owns that workflow. If they are curious about AI, give them this quarter to redesign it. If they aren't, the rollout for that team will fail no matter what tool you buy. Move the workflow or pair them with a redesigner. 3. Set a measurable target. "We currently spend 6 hours a week on this. By the end of Q3 we want to be at 90 minutes." Hours are easier to track than ROI and they correlate with margin within two quarters. That is the quiet version of Cuban's advice. Don't sprint to avoid bankruptcy. Walk past the firms that are sprinting and the firms that are standing still. Both are losing in different ways. ## FAQ **Is Mark Cuban wrong about AI?** Not fully. His underlying point, which is that AI adoption will separate stronger and weaker firms over the next decade, is broadly right. The framing of bankruptcy as the consequence is wrong for SMBs. The realistic consequence is plateau and gradual margin compression, which is harder to see and slower to fix. **What's the difference between a winner and a loser in Cuban's frame?** Cuban implies the difference is whether you adopted AI fast enough. The Gallup 2026 data and what I see across 120+ projects say the difference is whether a manager redesigned a workflow around the AI, not whether the company bought the tool. **Should I worry about going out of business if I haven't started with AI?** Probably not in the short term. You should worry about losing 1 to 3% margin a year to faster competitors. That compounds over three to five years and turns into a real number on the P&L without ever showing up as a crisis. **What's a realistic first AI project for a 30-person SMB?** The workflow with the highest manual hours per week. For most professional services firms that's proposal writing or client reporting. For most product firms it's customer support tier one or onboarding. Pick the one where redesigning it would free up the most useful capacity, not the one with the flashiest demo. **How long should the first project take?** Six weeks of elapsed time, with maybe 30 to 40 hours of total work spread across the manager, the team, and an outside hand. Anything longer and the budget will quietly inflate. Anything shorter and you probably haven't redesigned the workflow, you've just bolted a chatbot onto it. If you want a structured way to find your one workflow worth redesigning, the AI Quick-Wins Checklist runs through the questions I ask in the first 30 minutes of every audit. It's the 10-minute version. https://richardbatt.co.uk/quick-wins --- ## More about Richard Batt Richard Batt is an AI implementation specialist who helps businesses deploy working AI automation in days, not months. 120+ projects across 15+ industries. ### Key pages - [Home](https://richardbatt.com/) - [About Richard](https://richardbatt.com/about) - [Blog](https://richardbatt.com/blog) - [Contact](https://richardbatt.com/contact) - [Subscribe](https://richardbatt.com/subscribe) ### Contact - Email: richard@richardbatt.com - Location: Middlesbrough, UK (working globally) - Website: https://richardbatt.com