Waze Update Bugs: What 1,500 Complaints Reveal (2026)

Waze is not a failing app. It carries 4.31 stars from 8,905,674 Google Play ratings and 4.84 from 3,185,654 App Store ratings, and 500 million installs. Most drivers who open it get where they are going.

But 1,610,957 of those Play ratings, 18.1% of every rating the app has ever received, are 1-3 stars, and the recent ones are unusually specific about what changed and when. So on August 17, 2026 we pulled every negative review we could reach from both stores, tagged each one by version and storefront, and looked for complaints that move with a build number rather than with the weather.

Two of them do. One is a change on Android whose first reports are one week old and appears in five languages. The other is a bug on iPhone that was severe in June, and is now largely gone. Everything else in this set is noise, opinion, or something we cannot check.

What We Did, and What These Numbers Can and Cannot Show

The set is 1,500 negative reviews, dated June 1 to August 16, 2026:

"Negative" means 1-3 stars with at least 80 characters of text, so one-word ratings are excluded. Reviews were deduplicated by store ID before counting. Every review carries the app version the reviewer had installed when they wrote it, which is what makes the rest of this article possible.

Three limits matter, and they shape every claim below:

Raw monthly counts are a sampling artifact, not a trend. Both stores return newest-first with a hard cap per storefront, so the number of reviews we can retrieve per month is driven by how far back each storefront's cap reaches, not by how many people complained. In this set the Play window returns 116 reviews for June and 420 for August. That does not mean complaints tripled. It means the retrieval window is denser near the present. So we never compare raw counts across months. We only compare shares: what percentage of one month's negative reviews, or one version's negative reviews, mention a given thing. Shares are immune to that bias.

Version labels are the reviewer's installed build, not necessarily the build that caused the problem. Someone can install 5.22 and then complain about something that has annoyed them since 5.19. This blurs the signal in one direction only: it makes version-specific problems look less version-specific than they are. When a complaint still concentrates in one build after that blurring, the concentration is real.

We cannot verify a route. More on that below, because it is the single biggest category here.

Finding 1: On Android, Build 5.22.0.3 Took Away the Back Button Exit

This is the cleanest result in the set.

Seventeen reviews in the Play window say the same thing: the back button no longer closes Waze, it just minimizes it, and the app keeps running in the background. All seventeen were written in August 2026. Fifteen of them are on one build, 5.22.0.3.

The share comparison is what makes it conclusive:

A fiftyfold jump in one build, from a version with nearly four times the sample size, is not sampling noise. The dates back this up: the reports arrive on August 10, 11, 12, 13, 14, 15 and 16, every single day, rather than clustering on one bad afternoon.

They are also not one country's problem. The seventeen reviews come from five storefront locales: nine US English, three Hebrew, two Mexican Spanish, two Brazilian Portuguese, one German. Reviewers describe the identical mechanic in each:

I used to be able to press the back button and waze would shut down but that no longer works and now it is a pain shutting it down.

Latest update removed app shutdown via back button on Android. Now I have to do more taps to shutdown the app, contributing to more distracted driving.

Back button does not close Waze but minimizes it. No easy way to find how to change setting.

Portuguese

a última atualização não permite mais fechar o app com o botão voltar. agora tem que encerrar a rota, abrir o menu e clicar em desligar.

German

Lässt sich nach letztem Update nicht mehr beenden, ohne die App über den Taskmanager zu killen. Läuft also unbemerkt weiter und frisst Akku.

Hebrew

בעדכון האחרון לא ניתן לכבות את היישום בלחיצה עם החץ אחורה והיישום נשאר פתוח וצורך סוללה סתם.

Spanish

no deja cerrar la app y eso causa más consumo de batería.

Notice what the German, Hebrew and Spanish reviewers independently add: the app is still running, and it is eating battery. That is the second-order complaint, and it is why this is not purely cosmetic.

One reviewer put the open question better than we could:

Waze no longer exits with the back button. Not sure if this is a bug or a deliberate design choice in a recent update.

That distinction matters and we cannot settle it. A navigation app that survives a back press is a defensible design decision, since drivers hit back by accident constantly. Waze has not published a note about it that we can point to, so the reviews are the only public record right now. What we can say is that whatever the intent, it shipped without an obvious in-app way to opt out, which is what most of these seventeen are actually angry about.

The iPhone side is the control group. Across 800 negative App Store reviews in the same window, zero describe this problem. That is exactly what you would expect, because iOS has no system back button to change. When a complaint appears on one platform, in one build, and is mechanically impossible on the other, you are looking at a real regression rather than a mood.

You can read the raw feed yourself on the Waze Android complaints page, which updates as new reviews land.

Finding 2: The CarPlay Blank Map Wave Was Real, and It Is Mostly Over

The largest single theme in the iPhone set is not routing. It is CarPlay: 156 of 800 negative App Store reviews mention it, 19.5%, and 84 of those are one star.

But the interesting part is the shape over time. Here is the share of each month's negative iOS reviews that mention CarPlay:

In June, nearly half of everything negative said on the App Store was about CarPlay. By July it was one in twenty.

The version breakdown says the same thing, and points at where it started:

Six builds, one clean arc: it breaks around 5.19.1.0, stays broken through 5.19.95.0, starts recovering in 5.20.0.1, and is back to background level by 5.20.1.0.

The symptom is consistent enough to be diagnosable. Of the 156 CarPlay reviews, 31 explicitly describe a blank, white or black screen, 37 mention audio, 18 mention freezing or crashing, and 12 mention disconnects. The blank-map ones are the most precise:

The map rendering screen goes entirely blank (sometimes black, sometimes white). Navigation audio and turn-by-turn arrows still work in the corner, but the actual map layer fails to load or draw visually on the vehicle’s screen.

One of the recent updates has stopped it working in CarPlay. Just presented with a white blank map and relevant touch buttons round the side.

Over the last couple of weeks the car screen is going blank and unable to do anything about it - in both mine and husbands cars using different phones in both.

After last update maps disappears on car play after clicking on alternate routes.

Dutch

Na de laatste update, geen plattegronden meer tijdens het rijen met Apple CarPlay.

German

Neuestes Feature bei Apple car play: manchmal beim einstellen der Route wird die Karte weiß. Anzeige wohin mit Pfeilen gibt es noch aber die Karte ist einfach weiß.

Portuguese

Ao recalcular a rota, o mapa some da tela. Por favor corrijam. É no Apple CarPlay

Spanish

Desde hace unas semanas cuando pulsas la tecla de las rutas la pantalla se queda en blanco y tiene que cerrar y volver abrir la aplicación.

Italian

Dopo il penultimo aggiornamento talvolta con CarPlay non disegna la mappa. Con l’ultima praticamente non si vede più nulla, se non le svolte nei messaggi baloon

Read those together and a mechanism emerges that no single review states: the map layer stops drawing while the rest of the projection keeps working. Turn arrows, voice guidance and the surrounding buttons are all still there. Several reviewers also report it triggering specifically when they tap for alternate routes, and that force-quitting the app or reconnecting restores it until next time.

It was not one car brand or one country. The 121 June reports come from twelve storefronts: Brazil 26, Italy 17, Mexico 15, UK 14, Canada 13, Netherlands 10, Spain 9, Australia 7, Germany 5, Poland 2, Israel 2, Indonesia 1.

And it got fixed. We are not inferring that only from the falling share. A reviewer said so directly in late June, on 5.20.0.1:

Waze has finally fixed the CarPlay issue after the latest update, which is a big relief.

There is still a tail. Fourteen August reviews on 5.22.0.1 mention CarPlay, and at least one describes the blank map persisting for months, along with a support process that did not have a category for it:

For several months I keep getting a blank map when starting up Waze on Apple CarPlay. The workaround is to view Waze in Apple screen mode but this means having to look at a very small map.

So the honest read is: a serious, widespread CarPlay rendering bug ran from roughly late May through June 2026, was substantially fixed by the 5.20.1.0 build, and a residual version of it still affects some drivers. If you are on iPhone and your Waze map is blank in the car today, you are in a much smaller group than you would have been eight weeks ago.

Android Auto did not do the same thing, which is the check that keeps this from being blamed on something Waze-wide. The Android Auto share of Play negatives went 12.9% in June, 7.3% in July, 4.0% in August: elevated, drifting down, but never anywhere near CarPlay's 44%. Whatever broke was on the CarPlay projection path specifically.

Finding 3: The Biggest Recurring Theme Is Route Quality, and We Cannot Verify One Word of It

Across both stores, the most common thing negative reviewers talk about is the route itself: 63 Play reviews (9.0%) and 107 App Store reviews (13.4%) describe the routing as wrong, worse, longer, or leading them somewhere they did not want to go.

We are reporting that number and then leaving it alone, on purpose. A review saying the route was bad is not evidence that the route was bad. We were not in the car, we do not know what the traffic was doing, we cannot see whether the road was actually closed, and we have no way to distinguish a genuine routing error from a driver who preferred a different way. Counting these tells you what frustrates people. It does not tell you whether Waze is right.

What the reviews do give us is texture, and the texture is consistent across languages:

French

Une horreur, il te fait passer sur un circuit de tramway à Fribourg-en-Brisgau, interdit aux voitures.

German

Die Ausweichrouten sind die, die alle fahren und man steht dann dort mit den anderen im Stau.

Polish

aplikacja działa z małym opóżnieniem powiadamia na skrzyżowaniu zamiast 50 metrów wcześniej.

Turkish

durup dururken bizi saçma sapan bir yola saptirdi , üstüne birde feribot la geçip 23 Euro ödemek zorunda kaldım.

The German complaint is the one worth pausing on, because it is a criticism of the product working as designed rather than failing: if a traffic-aware app sends everyone down the same detour, the detour fills up. That is a known limitation of crowd-sourced rerouting, not a bug, and no amount of review data can resolve it.

Twenty-four of the routing reviews specifically say routes have gotten longer or slower, and seven describe being routed into traffic rather than around it. Both numbers are too small, and too subjective, to support a claim that Waze's routing has actually degraded. We are noting them so that nobody reads the silence as a clean bill of health either.

Forty-two Play reviews and 40 App Store reviews name Google Maps directly, usually as the thing they switched to or are comparing against. If you want the category view rather than one app, we ranked the field in Google Maps vs Apple Maps vs Waze.

Finding 4: Sound and Speed Alerts Split Along Platform Lines

This is the one cluster where Android reviewers complain more than iPhone reviewers: 67 Play reviews (9.6%) versus 53 App Store reviews (6.6%) describe sound, voice guidance or audio alerts failing.

Inside that, the sharpest sub-complaint is speed alerts going silent while staying visible on screen. Twenty-six Play reviews and 16 App Store reviews mention speed limits or speed cameras, and the recurring shape is that the icon appears but the beep does not:

The audio alert for going over the speed limit has stopped working. I've tried everything I can think of to get it working again, all other audio notifications are working e.g. speed bumps, potholes.

Italian

la segnalazione con un bip sonoro al superamento del limite impostato, ora non funziona più, cambia l'aspetto, ma il segnale acustico è sparito

Polish

Brak dźwiękowego komunikatu o odcinkowym pomiarze prędkości, przy rozpoczęciu i zakończeniu pomiaru.

Turkish

sesli yönlendirmeler çok zayıf hatta şehir içerisinde sesli yönlendirmeleri hiç yapmıyor o trafik içinde ekrana bakmak zorunda kalıyorsunuz

An iPhone reviewer in the UK describes the same asymmetry and why it matters at speed:

speed cameras are shown on the map, but I no longer receive any voice alerts or notifications for them. I only get the visual icon

A visual-only warning is close to useless in the exact situation the warning exists for, which is why these reviews are angrier than their word count suggests. This cluster is not version-locked the way the first two findings are. It appears across builds, so we are not going to claim an update caused it.

Finding 5: The Rest of the Tail

Everything else in the set is smaller. Listed honestly, with both stores side by side:

Two entries in that table deserve a note.

Crash and freeze is more than twice as common on iPhone (9.3% versus 4.4%). Part of that is definitional, since the CarPlay reviews often describe freezing too and get counted in both rows. We are flagging the overlap rather than presenting the gap as clean.

Ads are almost a non-issue in this data. Waze is ad-supported with no in-app purchases, which usually produces a large ad complaint cluster. Here it is 11 Play reviews and 9 App Store reviews out of 1,500. Whatever drivers are unhappy about, it is not the ad load.

Address search is the quiet one worth watching. Forty-four reviews on each store describe the search box failing to find places, freezing when typing, or rejecting valid addresses:

Hebrew

הוויז לא נותן לי להזין כתובת ופשוט נתקע המסך!!! מעצבן

The GPS row is separate and steadier, at 4.4% and 4.1%. These reviewers are not describing a wrong route, they are describing the app not getting a fix at all, which reinstalling does not solve:

Indonesian

setiap kali digunakan tidak terhubung dgn gps padahal gps aktif, dan coba delete aplikasi terus dowload lagi sama aja hasil, knp ini ya

Seven Play reviews mention Gemini by name, mostly around voice reporting or Android Auto handoff, and one links a specific regression to a specific date:

Italian

Con l' aggiornamento del 30 giugno, su android auto Waze non si collega alla navigazione automaticamente se lo si chiede a Gemini. Prima lo faceva.

Seven reviews is not a finding. It is a thread worth pulling on the next pass, and we would rather say that than round it up into one.

The Two Stores Disagree by Half a Star, and That Is Normal

Waze scores 4.31 on Google Play and 4.84 on the App Store. That is a wide gap for the same app, and it is tempting to read it as "the Android build is worse."

Do not. The two stores collect ratings in structurally different ways, prompt at different moments, and have accumulated their totals over different histories. We wrote up the mechanics separately in Google Play vs App Store reviews, and the short version is that a half-star gap between the stores is closer to the norm than the exception for a large app.

What is more useful than the gap is where each store's complaints concentrate, which this article is really about: iPhone reviewers are overwhelmingly talking about the in-car projection layer, and Android reviewers are talking about audio and, this month, an app that will not close.

For the code-computed trust summary rather than our reading, see is Waze legit, and the iOS review page carries the App Store side of the same feed.

What You Can Actually Do Today

Sorted by how many reviews suggest each one worked, not by what a support page would tell you:

• If Waze will not close on Android: end the route first, then use the app's own menu and its power-off option. Multiple reviewers describe finding it there after the back button stopped working. It is more taps than before, which is the complaint, but it does shut the app down rather than leaving it running.

• If your CarPlay map is blank: force-quit Waze on the phone and reconnect. This is the single most repeated workaround in the June set. Several reviewers also report that toggling the CarPlay split-screen dashboard view forces the map to redraw. If you are on a build older than 5.20.1.0, updating is the actual fix.

• If speed alerts are silent: check that the alert sound is enabled separately from voice guidance. Reviewers repeatedly report that turn instructions still work while the speed beep does not, which suggests two independent settings rather than one broken audio path.

• If the app is draining battery: this is downstream of problem 1 on Android right now. An app that never closes keeps holding location. Confirm it is actually shut rather than minimized.

• If routing is the problem: nothing in this data supports a fix, and we are not going to invent one. Try the alternate-route list before you set off, and if it consistently sends you somewhere wrong in your area, the report tools inside the app are the only mechanism that feeds back into the map.

If you are weighing a switch, the alternatives to Waze list is generated from the same-category leaderboard rather than from our opinion.

What This Set Does Not Prove

Stated plainly, because a data article that only lists its strengths is doing PR:

• It is not a sample of Waze users. It is a sample of people annoyed enough to write 80 or more characters. A 4.31 average on nine million ratings means the overwhelming majority never appear here at all.
• It cannot count anything a reviewer did not name. If ten thousand drivers hit the blank CarPlay map and shrugged, they are invisible to this method.
• The version attribution is one-directional. It can only understate how version-specific a problem is, never overstate it.
• Two findings do not make a pattern. We found exactly two complaints in 1,500 that move with a build number. The rest of the set is not evidence that everything else is fine, only that nothing else in it is measurable this way.
• We have no access to Waze's own telemetry, no comment from the company, and no way to see fixes that shipped without anyone mentioning them.

Bottom Line

Two things in this data are real and datable. On Android, build 5.22.0.3 changed what the back button does, the reports started on August 10, 2026, they arrive daily in five languages, and there is no equivalent on iPhone because there is no equivalent button. On iPhone, a CarPlay map-rendering bug dominated June, accounting for 44% of everything negative said that month, and was substantially fixed by build 5.20.1.0.

Everything else here is either unverifiable by design, like route quality, or too small to carry weight, like the Gemini handoff. That is a less dramatic conclusion than "Waze is broken," and it is the one the reviews actually support.

For the live feed instead of this snapshot, the Waze complaints page shows the newest 1-3 star reviews as they arrive.

Related reading: Google Maps vs Apple Maps vs Waze ranks the navigation category by complaint volume, Public Transit Apps Ranked covers the same problem for people who are not driving, and The Worst Rated Apps on Both Stores puts a 4.31 average in context.