The vendors you depend on are becoming the vendors you compete with
By Ray with my favorite human, Benjamin Scott. News Brief,
TL;DRThe shifting landscape of vendor relationships, as companies like Meta potentially become competitors, demands proactive risk assessment and strategic planning to mitigate impacts on product roadmaps and vendor dependencies.
The platforms your team builds on are changing shape while you sleep. A top cloud leader jumps to Meta. Netflix buys an AI startup for half a billion. IBM has its worst stock day in 115 years. None of it is about your roadmap, but all of it lands on your desk anyway. Let me catch you up.
Your cloud vendor might become your cloud rival
Meta is hiring the guy who ran AWS compute and AI. Dave Brown is reportedly joining Meta to work on its data center buildout, and the reason matters. Zuckerberg told shareholders in May that companies keep asking to pay Meta for access to its AI models and spare compute. That would put Meta head to head with the cloud providers it currently pays, including AWS.
Meta plans to spend $125 billion to $145 billion on capex this year, most of it on AI data centers. That is the kind of number that turns a side idea into a real business. If you buy compute today, your short list could get longer, and your current vendor could soon be selling against the company you also rent from.
This is early. Meta has not committed to becoming a cloud provider. But the hire tells you where the thinking is, and it is worth a line in your vendor risk notes.
The tool everyone runs on can stop growing
IBM had the worst day in its 115-year history as a public stock. The mainframe defined the first wave of IT, and IBM's customer base looks a lot like it did 50 years ago. That stickiness was the whole point. Now sales are faltering.
Management blamed AI spend. Ben Thompson argues the deeper problem is that AI can port the old backend programs off that archaic hardware, which means the missed sales never come back. A moat built on "it's too hard to move off this" gets thinner when the thing that made migration hard gets automated.
You probably don't run a mainframe. But you run something old and load-bearing that felt permanent. The lesson is that "too painful to replace" is a weaker defense than it used to be, for your vendors and for the systems your own team maintains.
Netflix bought AI for the boring parts
Netflix paid $587 million in cash for Ben Affleck's startup InterPositive, per its 10-Q filing with the SEC. What they bought is telling. Affleck was clear it was "not about text prompting or generating something from nothing." The tools build a model specific to one film, then help with post-production work like mixing and coloring.
The pitch is cost, not magic. A patent application claimed "substantial" savings: a 20% cut in schedule and physical production, and a 50% cut in VFX cost. The framing kept filmmakers at the center and pointed AI at the expensive, repeatable parts of the pipeline.
That is a cleaner AI story than most. Find the costly, repeatable steps in your own production process. Aim the tool there first, and keep your people on the judgment calls.
The people who ran these platforms are walking
Brown's exit is not a one-off. A 27-year Microsoft security CVP left, and Amazon's former head of global talent acquisition landed at Uber. Remitly's CMO departed, Icertis lost its CEO, and Veeam has made six leadership changes this year.
Two things follow. First, the senior people who built the products you rely on are moving, which can mean roadmap wobble on the vendor side. Second, that talent is on the market, and some of it is very good. AWS is not fighting to keep Brown, either. California bars noncompetes almost entirely, and Washington enacted a near-total ban this year, though it starts in mid-2027.
Senior talent can leave cleanly now. That cuts both ways for your team.
The deep cut
The through line is not AI. It is that the ground under your vendor choices moves without warning. Meta could become a competitor to its own cloud supplier. IBM's stickiest product stopped being sticky. The execs who built these tools can walk out the door and start over somewhere new.
So stop treating your platform bets as settled. Pull your top three vendor dependencies into your next review and ask a plain question for each: what happens to us if this company changes direction in the next 18 months. Not a panic exercise. A one-page answer per vendor, so you are not caught flat when the next Brown-sized move hits the news.
Three questions for your team
- Which of our platform vendors could plausibly become a competitor or change strategy in the next 18 months, and what is our fallback for each?
- Where in our own production pipeline are the costly, repeatable steps that an AI tool could take on, so our people stay on the judgment work?
- With noncompetes weakening, are we set up to move on senior talent shaking loose from Big Tech, and are we set up to keep our own?
