Everything About AI
NextFin News - Verisure Plc is bringing in outside technology and investing more in training artificial intelligence models on its own data as the Swedish home-security business seeks to reassure investors that it can keep up with new competition. Chief Executive Officer Austin Lally said the company is drawing data from its 6.4 million subscribers to train AI models to better distinguish genuine security threats from false alarms, technology already deployed in its latest detectors to identify people and animals and improve monitoring in difficult outdoor conditions.
The push comes at a delicate moment for a company that has been punished by the market for the same theme it is now embracing. Since debuting in Stockholm in October 2025, Verisure has lost more than a quarter of its market value as investors grew concerned that artificial intelligence could disrupt the professionally monitored security model that generates most of its revenue. In February, shares tumbled more than 17 percent intraday — the worst session since the IPO — after quarterly results failed to calm those fears, with the stock touching €10.35, well below the €13.25 offering price.
Verisure's answer is to weaponize the one asset a startup cannot easily replicate: years of alarm-trigger data from millions of homes. But the strategy rests on a partnership with Origin AI whose owner is now ADT Inc., the very competitor Verisure is trying to hold at bay. That tension — between a defensible data advantage and a technology supply line controlled by a rival — is the real story behind the AI pivot.
The Disruption Fear That Broke the IPO Glow
Verisure's listing was supposed to be a triumph. The offering raised about €3.2 billion, the largest European debut in three years, and shares closed the first day at €16 — 21 percent above the €13.25 offering price. The company was the largest IPO in Sweden since 2000 and accounted for more than half of the $7.23 billion raised from IPOs on the Stockholm exchange in 2025.
Nine months later, the mood had soured. The concern is not that Verisure lacks customers — it added 219,000 in the second quarter alone, bringing its base to 6.377 million, up 9.4 percent year over year.
In the second quarter we grew our portfolio by 9.4% to ~6.4m customers demonstrating resilience in a more uncertain consumer backdrop.
The concern is what happens to the economics of a business built on professional monitoring when AI makes alarms smarter, cheaper, and potentially self-sufficient.
The market delivered its verdict in February. After quarterly earnings, shares dropped the most on record, falling more than 17 percent at 10:57 a.m. in Stockholm and as low as €10.35. The sell-off was not about the numbers — annualized recurring revenue rose 11.9 percent to €3.62 billion, adjusted EBIT climbed 13.9 percent to €269 million, and free cash flow turned positive at €56 million. It was about the narrative: in an AI-driven market, is the human-in-the-loop monitoring model an asset or a cost center waiting to be automated away?
That is the question Lally is now trying to answer with technology and contracts rather than earnings calls.
The Two-Pronged AI Defense: Data Moat Plus Contractual Lock-In
Verisure's AI strategy has two distinct components, and they are not interchangeable.
The first is proprietary. The company is using alarm-trigger data from its 6.4 million subscribers to train models that reduce false alarms — the single largest operational cost and customer-friction point in monitored security. A detector that can tell a person from a pet, or a break-in from a swaying branch, means fewer unnecessary dispatches, lower operating costs, and fewer annoyed customers canceling their subscriptions. This is not theoretical: Lally said the technology is already in the company's latest detectors, identifying people and animals and improving monitoring in difficult outdoor conditions.
The second is contractual. In a long-term commercial agreement announced earlier this year, ADT, Origin AI and Verisure agreed to scale AI-driven security technology. Verisure has used Origin AI's sensing technology for five years and has been a minority investor since 2023. The deal grants Verisure long-term rights to certain AI sensing use cases, including age-technology applications, with provisions to extend those rights beyond the current term, and — critically — defined exclusivity in professionally monitored security across Europe and Latin America. The agreement is valued at $30 million over five years for development services, plus a per-household activation fee aligned with Verisure's existing arrangement with Origin AI.
The structure matters. Verisure is not trying to build foundational AI models from scratch — a capital-intensive race it would likely lose against well-funded technology giants. Instead, it is doing what incumbents do best: securing exclusive distribution rights to best-in-class technology while layering its proprietary data on top to differentiate the end product. The data improves the model; the contract keeps the model out of competitors' hands in its core markets.
The Complication: Your Partner's Technology Now Belongs to Your Rival
There is a wrinkle in the partnership strategy that investors should not gloss over. On February 24, ADT Inc. acquired Origin Wireless, the company behind Origin AI's Wi-Fi sensing technology, for $170 million in cash. Origin's technology can distinguish between human and non-human motion — analyzing whether activity comes from a person, a dog, or a mechanical object such as a robotic vacuum — using Wi-Fi signals from ordinary devices like laptops and smart light bulbs, without cameras or additional sensors.
So Verisure's AI edge depends on a technology stack now owned by ADT, one of the world's largest security companies and a direct competitor outside Verisure's exclusive territories.
The exclusivity clause is Verisure's shield: in Europe and Latin America, ADT cannot deploy Origin's sensing technology in professionally monitored security for Verisure's rivals. But exclusivity is geographically bounded and use-case specific. It does not prevent ADT from using Origin's technology in North America, in non-professional DIY systems, or in adjacent applications like age-tech and ambient assisted living — markets where Verisure has less presence and where the technology could be perfected before expanding into contested territory.
This creates an asymmetry. Verisure gains a protected runway in its home markets; ADT gains a technology platform with global optionality. Whether that asymmetry matters depends on one thing: how quickly AI-enabled self-monitoring becomes good enough to substitute for professional monitoring altogether.
Cyclical Sentiment Versus Structural Change: What Is Actually at Risk
It is useful to separate two things the market has conflated.
The stock decline is cyclical. Verisure's fundamentals have not broken: revenue is growing at a double-digit pace, margins are expanding — adjusted EBIT margin reached 26.3 percent in the second quarter, up 107 basis points year over year in constant currency — customer acquisition is healthy, and attrition is stable at 7.5 percent annualized. The de-rating reflects a sentiment shift, investors repricing the entire monitored-security category on the possibility that AI changes the rules. Sentiment mean-reverts when fundamentals hold, and Verisure's numbers suggest the business is not deteriorating.
The technological shift, however, is structural. AI is not a passing cost cycle; it is a permanent change in what security hardware can do. Motion detection that classifies rather than merely detects, Wi-Fi sensing that requires no new hardware, and camera-free monitoring all lower the barrier to entry and raise the threat of disintermediation. This will not revert on its own. An incumbent that ignores it does not get a reprieve — it gets replaced.
The critical analytical question is whether AI is substitutive or complementary to professional monitoring. If AI enables homeowners to monitor themselves reliably, Verisure's subscription model is structurally threatened. If AI makes professional monitoring more accurate, cheaper to deliver, and harder for DIY systems to match, then AI is a margin-expanding tailwind — and Verisure's data advantage becomes a genuine moat.
The evidence so far points toward complementarity, for one reason: false alarms are the enemy of every security system, DIY or professional. AI that reduces false alarms makes any system better, but a professionally monitored service can compound that advantage with human verification and dispatch — something a self-monitored camera cannot do. Verisure said it supported customers in more than 104,000 incidents in the fourth quarter that required intervention by police, fire, ambulance, or a guard. The value is not in detecting motion; it is in knowing which motion matters and acting on it. AI improves the detection; the monitoring network captures the value.
The Counter-Thesis: Why the Moat Might Not Hold
The strongest case against Verisure's strategy is straightforward and comes down to control. ADT now owns the technology. Contracts can be renegotiated, exclusivity can be litigated or circumvented, and a well-capitalized owner has incentives to prioritize its own deployments over a partner's. If ADT decides to compete aggressively in Europe with its own branded AI monitoring — or if a new entrant develops comparable sensing without Origin's patents — Verisure's contractual shield thins quickly.
There is also the question of whether data scale actually wins in AI. The argument that millions of subscribers' worth of alarm data creates an unassailable model assumes that more labeled data produces meaningfully better performance. That has held true in some domains — speech recognition, translation — and less so in others, where architecture and compute matter more than marginal data increments. If the next generation of security AI is won on model architecture rather than labeled trigger data, Verisure's data moat is shallower than it appears.
Finally, there is the customer-behavior risk that no contract can address: if AI makes DIY systems reliable enough, customers may simply stop paying for professional monitoring. Verisure's entire valuation rests on the assumption that households and small businesses will keep paying a monthly fee for someone else to watch their alarms. That assumption has held for two decades. AI is the first technology that could genuinely test it.
The falsifying signal is specific: watch Verisure's annualized attrition and average revenue per user. The company reported attrition stable at 7.5 percent annualized and ARPU of €48.20, up 2.4 percent year over year, in the second quarter. If attrition rises sustainably above 8.5 percent while ARPU growth stalls below 1 percent — particularly in the AI-competitive European markets — the moat thesis is broken, regardless of how many partnerships the company announces.
What Comes Next
In the short term, the stock will remain sentiment-driven. Investors have been burned once on the AI narrative and will want to see proof, not promises. The next two or three quarterly reports matter more than the next partnership announcement — specifically whether ARR growth holds near the guided 10 percent excluding Mexico, whether adjusted EBIT margin stays above 26 percent, and whether the €10 million rebranding investment begins to convert into lower acquisition costs.
In the medium term, the test is commercial: do AI-enabled detectors reduce false alarms measurably, and does that reduction show up in lower operating costs and higher retention? Verisure has the data to answer this internally; investors will get pieces of it through attrition and margin disclosures.
In the long term, the question is structural: does AI consolidate the security industry around players who own both the technology and the monitoring relationship, or does it fragment the market by enabling cheap self-monitoring? Verisure is betting on the former. Its partnerships buy time; its data is supposed to buy defensibility.
Bottom Line
Verisure's AI pivot is a rational response to a real threat, but it is a defensive move, not an offensive one. The company is using partnerships to access technology it cannot build and contracts to keep that technology from reaching rivals in its backyard — while hoping its subscriber data makes the combined product better than anything a competitor can field. That is a credible strategy for an incumbent with scale. It is also a strategy that leaves the company dependent on a technology owner that is also its competitor, in an industry where the endgame — fully reliable self-monitoring — would make both of them obsolete.
The market has priced Verisure as an AI casualty. The fundamentals do not support that yet. But the burden of proof has shifted: after a more than 25 percent post-IPO decline, reassurance requires more than partnership announcements. It requires the data to show that AI is making Verisure's monitoring more valuable, not less.
Explore more exclusive insights at nextfin.ai.
