How to Turn On Scam Protection in Google Messages

Picture the most ordinary scam message of 2026. It arrives at 11:42 on a Tuesday, from a number you’ve never seen: “Your parcel could not be delivered. Reschedule here:” followed by a link that looks almost — almost — like a courier’s website. You’re half-distracted, your thumb hovers, and that’s the whole game. The scammer doesn’t need you to be careless. They just need you to be busy.

Google has been quietly building a safety net for exactly this moment inside its Messages app, and on 14 October 2026 it announced the next step: an opt-in enhanced protection mode for Google Messages, built around on-device and cloud AI that studies conversations with unknown senders and interrupts the classic scam patterns before you act on them. It is rolling out gradually over the coming weeks, it only works if you switch it on yourself, and — crucially — not every Android phone gets every piece of it.

This guide explains what the new protections actually do, which phones get which features, how to check whether they’re on your phone, and how to turn them on. Everything here is based on Google’s October announcement and the release notes for the October 2026 Google Play system updates. Where a detail depends on your device or region, we’ll say so plainly.

Why Text-Message Scams Needed a Smarter Response

The old defences haven’t kept up. Spam filters catch bulk junk well enough, but the scams that hurt people today are conversational: a “bank” warning about suspicious activity, a wrong-number text that slowly becomes a romance pitch, a fake toll-road notice with a deadline today. They don’t arrive in waves; they arrive one by one, tailored just enough to feel personal.

Google’s answer has been moving in layers. The Messages app already flags suspected spam, and the October 2026 Play services updates extended in-call scam protection — which blocks accessibility-permission requests during live calls, the classic “stay on the line while I walk you through installing this app” trick — to Android 15 and earlier. The new enhanced protection mode adds a deeper layer: instead of only matching known scam signatures, AI models evaluate the conversation itself.

There’s a useful mental model here. Think of Messages’ protections as three concentric rings: spam filtering (catches the obvious junk), link and call safeguards (interrupts risky moments), and now enhanced protection (actively analyses unknown conversations). This guide is mostly about the third ring — and the honest truth is that it’s the one where your phone, your region, and your patience with rollouts matter most.

What Enhanced Protection Mode Actually Includes

Google described the October announcement as a family of features, not a single switch, and that’s worth understanding before you go hunting through settings. According to the announcement and subsequent reporting, here’s the breakdown:

Scam Detection with Private AI Compute

This is the headline feature. When a conversation with an unknown sender starts looking like a scam — the patterns Google’s models were trained on, such as fake delivery notices, prize claims, or “urgent account action” lures — cloud-based Gemini models analyse the conversation and warn you. Google says this runs through its Private AI Compute infrastructure, its privacy-focused cloud system for AI processing — which is why the feature can use powerful cloud models rather than relying solely on your phone’s hardware. Availability: Pixel 6 and newer, Galaxy S24 series and newer, Galaxy Z Flip and Z Fold 6 and newer, and other devices with Google’s AICore, rolling out over the coming weeks.

Gemini Nano link safety checks

Separately, an on-device model — Gemini Nano — checks unfamiliar links against real-time web signals before you tap through. The advantage of doing it on-device is speed and privacy: the link never has to leave your phone to be evaluated. Availability: Pixel 10 and newer, Galaxy S26 series, Galaxy Z Flip 8 and Z Fold 8, and other select devices.

Link checkpoints

A redesigned, harder-to-miss warning that appears before you open a link from a number not in your contacts. This one has the broadest reach — reportedly Android 6 and newer — which makes it the feature most readers of this site will actually encounter first.

Scam alerts in chat notifications

Warnings about suspicious chats that appear right in the notification shade, previously a Pixel-only feature, are extending to phones from HONOR, Infinix, iQOO, itel, JOVI, Motorola, OPPO, OnePlus, realme, TECNO, vivo, and Xiaomi later this year. Timing varies by manufacturer, so this is a “watch for it” rather than a “switch it on today” feature on most non-Pixel phones.

Screen-share privacy and call-forwarding protection

Two smaller but genuinely useful additions arrived alongside the Messages announcement. Screen-share privacy blanks out login and payment fields when you share your screen in Chrome or Android WebView (Android 15 and newer, rolling out through October), which closes off the “share your screen so I can help you” banking-scam pattern. Call-forwarding protection blocks apps from quietly forwarding your calls and warns you about dial-pad forwarding codes — scheduled for Pixel and select devices in early December.

Protection	What it guards against	Who gets it
Scam Detection (Private AI Compute)	Conversational scams from unknown senders	Pixel 6+, Galaxy S24+, Z Flip/Fold 6+, other AICore devices
Gemini Nano link safety checks	Malicious links from unfamiliar numbers	Pixel 10+, Galaxy S26, Z Flip 8 / Z Fold 8, select devices
Link checkpoints	Accidental taps on unknown-sender links	Android 6 and newer — the widest coverage
Chat notification scam alerts	Suspicious chats spotted at a glance	Pixel now; HONOR, Motorola, OnePlus, OPPO, Samsung, vivo, Xiaomi and others later in 2026
Screen-share privacy	Exposed logins during screen sharing	Android 15 and newer
Call-forwarding protection	Silent call forwarding by malicious apps	Pixel and select devices, early December

One important framing note: device lists above come from Google’s announcement as reported in October 2026, and Google has a long habit of expanding or adjusting availability after announcements. Treat the table as the plan, not a contract.

How to Turn On Enhanced Protection in Google Messages

The single most important fact: the new mode is opt-in. Google will not switch it on for you. Since it’s rolling out gradually, the setting may not be on your phone yet — so the first half of this process is checking you’re ready, and the second half is the actual switch.

Step 1: Update everything first

Open the Play Store, tap your profile picture, go to Manage apps & device, and update Google Messages and Google Play services if updates are waiting. The scam-protection features ship through these components, not through full OS updates — our guide to the October 2026 Google Play system update explains why so much of Android’s protection now arrives this way. An outdated Messages app is the most common reason the new options don’t appear.

Step 2: Find the protection settings in Messages

Open Google Messages, tap your profile picture (or the three-dot menu, depending on your version), then Messages settings and look for a Protection & safety or similarly named section. Because the rollout is in progress, the exact label and placement can differ between versions — if you don’t see an “enhanced protection” option, your phone likely hasn’t received it yet rather than you looking in the wrong place.

Step 3: Switch on enhanced protection

When the option is present, toggle it on and confirm any prompt. That’s the whole action. There is no account to create, no subscription, nothing to pay for. If a prompt mentions that conversations with unknown senders may be analysed by Google’s AI systems, that’s the Private AI Compute feature described above — worth reading, but it’s the core of what you’re enabling.

Step 4: Turn on the protections you already have

While you’re in settings, make sure the existing safeguards are active too, since they work alongside the new mode:

• Spam protection: in Messages settings, confirm spam protection is enabled so suspected spam is filtered before it reaches you.
• Verified SMS / business verification: where available, this marks messages from verified businesses, making impersonation easier to spot.
• In-call scam protection: check your phone’s security settings (usually under Settings > Security & privacy) for scam or call-protection options — the October Play services update extended these to Android 15 and earlier.

If spam calls (not just texts) are the bigger problem in your life, our guide on blocking spam calls and texts on Android covers the manual filters and call-screening tools worth setting up alongside Google’s automated protections.

If You Don’t See the New Options Yet

Most people reading this in late October 2026 won’t see the enhanced protection toggle — and that’s expected, not broken. Google said the rollout happens “in the coming weeks,” and these staged rollouts typically reach Pixel phones first, then flagships from other makers, then everyone else. A few practical things to try:

• Check again after updating. Rollouts often unlock server-side, so the toggle can appear days after your app is technically new enough. If you updated Messages this week, check again next week.
• Confirm your region. The announcement named the US among the first markets for the October 13 rollout; Google hasn’t given a worldwide timetable. Features sometimes arrive in your country months later.
• Don’t confuse it with old settings. If you only see the older spam protection toggle, you have the previous generation — still useful, just not the new AI-driven mode.
• Never sideload a “new Messages” APK to get it early. Unofficial APKs from random sites are one of the most reliable ways to install actual malware while chasing a security feature. If it’s not in the Play Store yet, waiting is the safe move. For context on installing APKs safely when you genuinely need to, see our APK installation guide — but a staged Google rollout is not a reason to sideload.

Common Mistakes People Make With Scam Protection

Protection features create a false-comfort trap if you misunderstand what they do. A few mistakes worth naming explicitly:

• Treating “no warning” as “safe.” No AI model catches every scam. A message that triggers no alert is not verified — it’s just unflagged. If a message creates urgency (“act now,” “account suspended,” “you’ve won”), slow down regardless of what your phone says.
• Trusting the sender name. Scammers routinely spoof bank and courier names in the sender field. The new link checkpoints help precisely because they evaluate the link, not the name — but only for links, and only when the feature has reached you.
• Granting accessibility permissions during calls. The in-call protection added in the October update exists because scammers talk victims through granting these permissions live. If anyone on a call asks you to change a security setting, end the call first — a real bank will never ask you to do this.
• Assuming the feature covers every messaging app. These protections live in Google Messages and the system dialler. WhatsApp, Telegram, and third-party SMS apps have their own (weaker or stronger) safeguards. A phone fully protected in Messages is still unprotected in an app where these features don’t exist.
• Ignoring app permission hygiene. Scam protection is a net, not a wall. Reviewing which apps hold sensitive permissions remains one of the highest-value security habits on Android — our app permissions guide walks through the ones that matter most.

When Enhanced Protection Matters Most — and When It Doesn’t

Who benefits most? If you regularly receive texts from unknown numbers — delivery notifications, two-factor codes, marketplace buyers, job recruiters — you’re in the highest-exposure group, and the unknown-sender analysis is built for exactly your inbox. Older relatives who are newer to smartphones are the other obvious group; if you manage a parent’s phone, switching this on for them during your next visit is one of the highest-value favours you can do.

When does it matter less? If you almost exclusively message saved contacts and rarely tap links from texts, the link checkpoints and conversation analysis will rarely fire — which is fine; they’re insurance, not a daily tool. And if your phone is several years old and outside the AICore device list, temper expectations: you’ll likely get the link checkpoints (Android 6+) but not the full AI-driven conversation analysis.

One more honest caveat: turning this on does not replace the fundamentals. Keep your phone’s system updates installing, keep Play Protect active, and keep a healthy suspicion of any message that wants you to act fast. The AI is a second pair of eyes, not a substitute for your own.

Frequently Asked Questions

Is enhanced protection mode in Google Messages free?

Yes. Google has described it as a free, opt-in feature of Google Messages and Google Play services. There is no subscription, no payment, and no developer account involved.

Does it read all my messages?

According to Google’s announcement, the conversation analysis applies to messages from unknown senders and runs through Private AI Compute, which Google says is designed so message content isn’t stored or reused. Your chats with saved contacts aren’t the target of this analysis. If that privacy model matters to you, read Google’s own description of Private AI Compute before enabling the feature.

Will it work on my Samsung / Xiaomi / OnePlus phone?

It depends on the feature. Link checkpoints are slated for Android 6 and newer, so most phones should get those. The full AI conversation analysis needs AICore-capable hardware — Google named recent Pixel and Galaxy flagships at launch, with other AICore devices to follow. Chat-notification scam alerts are expanding to HONOR, Infinix, iQOO, itel, JOVI, Motorola, OPPO, OnePlus, realme, TECNO, vivo, and Xiaomi later in 2026, with timing varying by manufacturer.

I enabled it but haven’t seen any warnings. Is it working?

Probably, yes. The protections only act when something suspicious happens — a scammy message from an unknown number, a risky link tap, a screen-share with sensitive fields. No warnings during normal texting with people you know is the expected behaviour, not a sign of failure.

Can scammers get around it?

Some will, eventually. Scam tactics evolve against every defence — that’s the history of spam filtering in a sentence. Enhanced protection raises the bar meaningfully, especially against the high-volume template scams, but it doesn’t make you immune. The habits still matter: verify unexpected requests through a separate channel, never grant permissions during a call, and treat urgency as a red flag.

The Bottom Line

Google’s October 2026 announcement is best understood as Android’s scam defences growing a brain: from matching known-bad patterns to actually reading the conversation. The opt-in enhanced protection mode, the broader link checkpoints, and the screen-share and call-forwarding safeguards together close several of the gaps scammers have been exploiting — especially the live-call social engineering that the October Play services update also targeted.

The practical move is simple: update Messages and Play services, open the protection settings, and switch on everything available to you. Then check back in a few weeks if the new toggle hasn’t arrived yet. And keep the oldest advice in the book, because no update will ever replace it: if a message is rushing you, it’s probably not your friend.