OpenAI's roll-up bought ADT for $10 a share — the house just became an AI agent's execution layer

Thrive Holdings agreed on Oct. 29, 2026 to acquire ADT for $10.00 a share in cash, valuing the equity at approximately $7.6 billion and the enterprise at $15.3 billion. The price is 45.6% above ADT's unaffected close of $6.87 on Oct. 27, 2026 and 52.1% above its 30-day volume-weighted average price, and ADT's board approved unanimously. Financing commitments come solely from funds managed by affiliates of Apollo Global Management, with no financing condition attached, and closing is expected in the first quarter of 2027 subject to shareholder approval and expiry of the Hart-Scott-Rodino waiting period.

Do the arithmetic on ADT's own filings and the offer looks far less like an AI transaction than a services transaction. Trailing revenue through June 30, 2026 was about $5.16 billion and trailing operating earnings plus depreciation and amortization about $2.71 billion, so $15.3 billion of enterprise value is roughly 3.0x sales and 5.7x EBITDA. Rebuild the pre-bid enterprise value at the $6.87 unaffected price and the same math gives closer to 4.8x. The whole AI-and-agent narrative was therefore bought for about 0.9 turns of EBITDA, or roughly $2.4 billion of enterprise value.

The balance sheet explains why the multiple looks the way it does. ADT ended June 30 with $7.69 billion of total debt, $3.9 million of cash and about $7.68 billion of net debt, or roughly 2.8x trailing EBITDA, against $404 million of trailing interest expense and coverage of about 6.7x EBITDA to interest. Cash generation is the reason the equity is worth anything at all: trailing operating cash flow of $2.16 billion and $1.88 billion of free cash flow, with management guiding to roughly 30% growth in adjusted free cash flow including swaps for 2026. Apollo's commitment, and the absence of a financing condition, removes the financing-out risk that usually discounts take-private spreads.

The bottleneck in home artificial intelligence is not intelligence, it is verification and actuation. A camera that sees a shape cannot tell a curtain from a person; a sensor that trips cannot decide whether to call the police. ADT's operating stack — around 250 sales and service offices, UL-listed monitoring centers, about 12,200 employees, independent authorized dealers who install the equipment, and a 24/7 monitoring operation — turns monitoring into the execution layer for home AI agents. That is what a language system cannot rent, and it is what Thrive has just bought.

This is not a new playbook. Thrive's accounting platform processed more than 7,000 tax returns at 98% accuracy and cut preparation times by over 30%, while its IT platform sped up help-desk resolution by 36 times and doubled the number of custom AI agents deployed in a month. Applied to monitoring, the equivalent milestone is measurable and brutal: fewer false dispatches, shorter handle time, higher retention on a base that churns 13.1% a year. If that works, false-alarm reduction becomes the first testable number and everything else follows from it.

The upstream dependency most people miss is software. Alarm.com's filings show ADT represented more than 15% but no more than 20% of its revenue in each of 2023, 2024 and 2025, and that after ADT's master service agreement ended, ADT now pays Alarm.com a monthly per-subscriber royalty for licensed patents at a rate Alarm.com itself describes as lower than its SaaS pricing. So Alarm.com already earns a below-SaaS royalty from its biggest customer, and that customer just became an artificial-intelligence competitor with $7.6 billion of fresh equity.

Now the privacy and liability bill. The announcement is explicit that ADT's data protection and privacy policies will remain unchanged and that no new data sharing will occur without customer permission, and a Thrive partner framed the same commitment around privacy and cybersecurity. Yet ADT's own 10-K warns that uncertainty in the development and deployment of AI, including generative AI, could hurt the business, and separately warns that independent dealers cannot fully mitigate data-security and product-liability risk. If an AI-assisted triage decides not to dispatch, the liability lands on a monitored services company whose franchise was built on human judgment.

The read-across is not one trade, it is a stack, and the stack is splitting in two. One camp puts intelligence in the device and locks the data down; the other puts intelligence in the cloud and needs the household to permit it. Thrive has just bought the cloud camp's execution layer for 5.7x EBITDA, and the two largest consumer platforms are choosing the device camp. Upstream, LG Electronics and Schneider Electric capture the physical bill of materials either way — cameras, a video doorbell, a smart deadbolt, a thermostat and a temperature sensor designed for Apple's Home Collection, with on-device recognition of people, pets, vehicles and motion.

The commercial logic for the DIY push is defensive, and that is the part that should worry the hardware incumbents. ADT Blu puts standalone cameras at $69 and full packages from $249 on Amazon and Walmart — the two shelves where Ring and Blink already live — which splits the smart home into trusted-cloud and private-device camps. Every sale that lands on those shelves is a sale that teaches a household to self-monitor rather than to subscribe.

Over the next two quarters this is an event-driven situation with dated checkpoints, not a fundamental one. ADT's third-quarter results arrive Nov. 3, 2026 — the last public operating data before shareholders vote. The go-shop ends at 11:59 p.m. ET on Nov. 28, 2026, with the extension for qualifying bidders running to Dec. 8. Hart-Scott-Rodino clearance and the shareholder vote follow, closing is expected in the first quarter of 2027, and the stock leaves the NYSE at completion, at which point ADT's $0.22 quarterly dividend needs Thrive's consent rather than its board's.

Over one to three years, the questions are narrower and harder. Does an AI-assisted monitoring operation hold or grow recurring monthly revenue above $360 million while gross attrition falls back under 13%? Does the DIY push feed the monitored business or cannibalize it? Does Alarm.com keep the ADT royalty, and if not, does Alarm.com's roughly 15% customer concentration turn from an annuity into a write-down? And does Resideo, newly standalone at about 4.5x its guided standalone EBITDA, become the next asset in this same playbook? On all of these, the first three quarters of AI-assisted monitoring will decide whether the AI thesis was ever real.