AI App Report DACH 2026
What state are apps built with Lovable, Base44 or Bolt in? 3,728 apps with a connection to Germany, Austria and Switzerland, scanned from the outside. All figures free to quote.
decivo's AI App Radar DACH checked 3,728 discoverable AI-built apps with a DACH connection from the outside. 75% of the apps checked in a browser had serious accessibility violations, 13% contacted trackers before any consent, 11% had no recognisable link to a legal notice or privacy policy, 6% shipped a library with a known vulnerability. Run “2026-Q4-recall”, methodology 2026.37, decivo's own survey, figures under CC BY 4.0.
Key points
- 75% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (2,449 apps) had at least one serious accessibility violation in the fully loaded page (automated axe-core check, impact “serious” or “critical”).
- 13% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (420 apps) contacted trackers on first load before any consent was given.
- 11% of the 3,728 discoverable AI-built apps with a DACH connection (424 apps) had no recognisable link to a legal notice or a privacy policy.
- 6% of the 3,728 discoverable AI-built apps with a DACH connection (240 apps) shipped a front-end library in a version with a publicly known vulnerability.
- All figures are under CC BY 4.0 and available as JSON. Quoting with attribution is explicitly welcome.
Sentences to quote
You can use these sentences word for word. Each one names share, base and count. Please include the source line below.
75% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (2,449 apps) had at least one serious accessibility violation in the fully loaded page (automated axe-core check, impact “serious” or “critical”).
13% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (420 apps) contacted trackers on first load before any consent was given.
7% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (243 apps) set tracking cookies before any consent was given.
11% of the 3,728 discoverable AI-built apps with a DACH connection (424 apps) had no recognisable link to a legal notice or a privacy policy.
6% of the 3,728 discoverable AI-built apps with a DACH connection (240 apps) shipped a front-end library in a version with a publicly known vulnerability.
56% of the 3,728 discoverable AI-built apps with a DACH connection (2,089 apps) answered a deliberately non-existent address with “200 OK” instead of an error.
26% of the 3,728 discoverable AI-built apps with a DACH connection (984 apps) shipped hardly any visible text in the initial HTML; the content only appeared in the browser.
Privacy
Three findings of severity “high”. All three describe an observed state, not a guess.
- Trackers fire before consent
- Legal notice or privacy policy not linked
- Tracking cookies set before consent
More in the chapter Privacy in AI-built apps.
Security
One finding of severity “high” and five missing protective layers. The layers are missing almost everywhere, also outside AI apps; their share is no measure of severity.
- Shipped library with a known vulnerability
- No Content-Security-Policy
- No clickjacking protection
- DMARC in monitoring mode only
- No security.txt
- No CAA record
More in the chapter Security of AI-built apps.
Accessibility
- Serious accessibility violations in the fully loaded page
The most frequent finding of the run that can prevent a “Go”. More in the chapter Accessibility of AI-built apps.
Technology and discoverability
These findings decide whether search engines and AI search can read an app at all and how fast it loads.
- Non-existent address answers “200 OK”
- No canonical address
- Content only appears in the browser
- First page load above 4 MB
- More than 1.5 MB of JavaScript
- No caching directives
26% of the 3,728 discoverable AI-built apps with a DACH connection (984 apps) shipped hardly any visible text in the initial HTML; the content only appeared in the browser. Anything that only reads HTML sees an almost empty page for these apps. That affects many crawlers used by AI search.
Our overall verdict
Every app gets a verdict. “Go” means: no finding of severity “high” or “critical”, fewer than three medium findings in the same area and no medium defect in security or privacy. Everything else counts as needing rework. “Stop” only applies to a critical finding confirmed from the outside.
- Needs rework (no “Go”)
- No urgent finding (“Go”)
“Stop” went to 17 of 3,728 apps. Sample: 3,728 reachable apps, of which 81% (3,030 of 3,728) were built with Lovable.
The overall verdict is our own threshold, not a single measured value. It also includes findings on technology and discoverability and rules that, in methodology 2026.37, already counted the mere mention of a service in the code. When measured again, the share without a “Go” may be lower. The individual findings are better suited for quoting.
Countries and builders
Share of apps with “Go”, by country and by builder.
- Germany: „Go“
- Switzerland: „Go“
- Austria: „Go“ (full inventory, not this run)
- Lovable (3,030 Apps): „Go“
- Base44 (383 Apps): „Go“
- Bolt (107 Apps): „Go“
- Wix (49 Apps): „Go“
More in the chapter Germany and Switzerland compared.
How it was measured
The targets come from public sources: the builders' directories, public certificate logs, the web archive, the open web index Common Crawl and the scan service urlscan.io. Only apps that are reachable and show a DACH connection on the start page enter the sample, such as a .de, .at or .ch domain, German page language or a link to a legal notice.
The scan runs from the outside and without logging in. It fills in no form and creates no account. For 88% of the apps the check in a real browser ran as well; findings only visible there count across those apps only. That is why some rows have a smaller base.
Only shares across groups of at least 20 apps are published: no hostname, no address, no finding that could be traced to a site. The full methodology is on the page of the run.
What the figures do not say
The radar measures what can be found. The 3,728 apps are not a random sample of all AI-built apps. The correct phrasing is “x% of the 3,728 discoverable AI-built apps with a DACH connection”, never “x% of all Lovable apps”.
The check runs from the outside and without logging in: what the page ships and what a browser observes while loading. The radar does not see database rules, roles or anything behind the login. A share says for how many apps a rule fired. It does not say whether the finding holds in each case, and it is not legal advice.
The report deliberately does not cite everything. At the time, for fonts, trackers, contacted services and findings in shipped code a mere mention also counted, for example in an embedded third-party script. Since methodology 2026.43 only what the page demonstrably loads or contains in its own code counts; exception: if the page ships no code of its own, the rules for database write calls and role comparisons check an embedded third-party script. These shares may be lower when measured again. Those shares are in the dataset, but not in this text.
The figures are decivo's own survey. decivo itself offers to fix such findings. They have not been independently verified so far; the dataset is open for that purpose.
The basis is the run “2026-Q4-recall”, closed on 8 November 2026. Across all 14 runs since September 2026 the radar has scanned 4,402 apps; the report uses the single run because it has one methodology and a fixed address.
Data and licence
All figures are under CC BY 4.0. You may reuse them, including in your own charts, if you name “decivo AI App Report DACH 2026” and link to this page. The raw figures as JSON contain every rule with count, base and a plain title.
The same dataset is also available as an unchanged copy at Zenodo (DOI 10.5281/zenodo.24420658) and Hugging Face. That keeps it retrievable and verifiable independently of this site.
This edition stays as it is. A new edition gets a new address. The latest state across what are now 4,402 apps is shown by the running radar on decivo.de.
The four chapters
- Privacy in AI-built apps
Trackers before consent, cookies, missing legal notice.
- Security of AI-built apps
Libraries with known vulnerabilities and missing protective layers.
- Accessibility of AI-built apps
Serious violations in the fully loaded page.
- Germany and Switzerland compared
Verdicts by country and by builder.
Sources
Every share comes from the radar run named below. Its fixed address shows exactly these figures; the legal sources are the statutes themselves.
Whether your app is part of these shares, the free scan shows in a few minutes: from the outside, without access, with evidence for each finding.
Check your app for freeQuestions about the report
May I quote the figures?
Yes. The figures are under CC BY 4.0. Name “decivo AI App Report DACH 2026” and link to this page.
How many AI-built apps have defects?
75% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (2,449 apps) had at least one serious accessibility violation in the fully loaded page (automated axe-core check, impact “serious” or “critical”). 13% of the 3,280 AI-built apps with a DACH connection that were checked in a browser (420 apps) contacted trackers on first load before any consent was given. 6% of the 3,728 discoverable AI-built apps with a DACH connection (240 apps) shipped a front-end library in a version with a publicly known vulnerability.
Do the figures apply to all Lovable apps?
No. They apply to the 3,728 apps the radar could find and that show a DACH connection. Apps on their own domain without traces of a builder are under-represented.
How often is the report published?
The radar runs continuously. The report summarises one closed run; every edition keeps its address and its figures.
Where does your app stand by comparison?
The decivo Scan checks your app from the outside, with the rules as they stand today, and names every finding with evidence.
30 minutes, free. You talk to us directly.