A hotel Wi-Fi network can now hand your Microsoft 365 access to an attacker without a single phishing email. Since at least June 2026, attackers have compromised captive portal gateways at hotels and conference venues and poisoned DNS for every guest. The fix is architectural: protect identity traffic before it ever reaches a network you do not control.
What Is the Gateway Campaign?
Threat researchers published details on July 23, 2026 of a campaign that targets the captive portal gateways guests use to get online. On a hotel Wi-Fi network, these appliances hand out IP addresses and act as the DNS resolver for everyone on the network. Therefore, an attacker who controls one gateway controls where every connected device goes. Compromised gateways sat in multiple US cities, India and Saudi Arabia, and traffic through them came from financial services, professional services, legal, healthcare, energy and retail organizations.
How Does a Hotel Wi-Fi Attack Reach Microsoft 365?
The chain runs in four steps. First, the attacker gains admin access to the gateway. Then the gateway, acting as the resolver, returns forged answers for Microsoft domains. Next, the traveler lands on attacker infrastructure that imitates Microsoft 365. Finally, in the cases researchers directly observed, the user approved a device code prompt, and the attacker received OAuth tokens with MFA already satisfied, with no password page at all.
Notably, hard-coded resolvers such as 8.8.8.8 did not help, because the lookups traveled unencrypted through the compromised gateway. In addition, the attackers resolved every query to malicious infrastructure rather than picking targets by keyword, which researchers read as a sign of a less careful operator.
How Did Attackers Get Into the Gateways?
Nobody has confirmed the entry point. Researchers assess, with low to medium confidence, that attackers used internet facing SSH, SNMP or web admin consoles together with weak or reused administrator passwords. That fits a familiar pattern, since hotel Wi-Fi gateways rarely get the patching and monitoring an enterprise firewall receives. However, visibility limits kept researchers from proving it, so treat this as a likely path, not a fact.
What Did Researchers Confirm, and What Do They Only Suspect?
The details below separate confirmed findings from assessments, so you can weigh the hotel Wi-Fi risk honestly.
| Finding | Status |
|---|---|
| Gateways compromised in US cities, India and Saudi Arabia | Confirmed |
| DNS poisoning of Microsoft domains by the gateway resolver | Confirmed |
| Four Microsoft-themed domains run by one actor | Confirmed, high confidence |
| Device code flow abuse yielding OAuth tokens | Directly observed, limited cases |
| WPAD proxy abuse | Attempted in about a third of cases, success not confirmed |
| Adversary-in-the-middle token theft | Not confirmed for this campaign |
| Entry through exposed admin consoles and weak passwords | Suspected, low to medium confidence |
| Tradecraft similar to APT28 | Assessment only, no shared infrastructure or code |
Above all, note the adversary-in-the-middle point. Many write ups describe session hijacking through a proxied login, and the MITRE ATT&CK adversary-in-the-middle technique explains that method well. Still, researchers list it only as tradecraft that overlaps with an earlier router campaign called FrostArmada, which authorities disrupted in April 2026, not as something they saw here.
Which Indicators Should You Search For?
Researchers published four Microsoft-themed domains, two hosting addresses and the DNS poisoning response address. Search DNS, proxy and Entra ID logs for them, especially for staff who used hotel Wi-Fi, and expect the actor to rotate infrastructure.
Why Does Standard MFA Fall Short Here?
Device code abuse turns MFA into a formality. The user approves a genuine Microsoft prompt, so the tokens the attacker receives already carry a completed MFA check. As a result, push approvals, SMS codes and one-time codes offer little protection once a hotel Wi-Fi gateway sits in the middle. Instead, phishing-resistant FIDO2 keys and passkeys bind sign-in to the real Microsoft domain, which an imitation site cannot satisfy.
How Do You Protect Travelers on Untrusted Networks?
Treat hotel Wi-Fi and every other public network as hostile, and enforce controls in this order.
- Always-on full-tunnel VPN: route all traffic, DNS included, to your own network, with no split tunnel for authentication, and block internet access until the tunnel is up
- Strict-mode encrypted DNS: use DNS over HTTPS or TLS with plaintext fallback disabled, because opportunistic mode falls back to the poisoned resolver
- Phishing-resistant MFA: move executives, administrators, legal, finance and frequent travelers to FIDO2 keys or passkeys first
- Block the device code flow: use a Microsoft Entra ID Conditional Access policy for users with no business need
- Disable WPAD: turn off automatic proxy discovery where you do not need it, or restrict PAC files to internal hosts
- Harden conditional access: require compliant devices and flag unfamiliar locations and risky sign-ins
Microsoft documents exactly how to block authentication flows with Conditional Access, including the device code flow, so that control can go live today.
What Should Security Teams Check After Employee Travel?
Start with identity logs, because endpoint tools will see nothing. Review Microsoft Entra ID for device code sign-ins, new OAuth grants, unfamiliar IP addresses and sessions created during or right after a trip. Then search DNS and proxy logs for the published domains, and ask travelers to report odd Microsoft prompts after hotel, conference or airport stays. Our 2026 Mid Year Threat Pattern Report tracked this same hotel Wi-Fi tradecraft among the year’s nation state patterns, so it deserves a standing place in your threat model.
How Can Digital Warfare Help?
A travel policy only works if the controls behind it hold. Our team runs penetration testing and vulnerability assessments that test whether your VPN, DNS and Conditional Access settings actually protect identity on hotel Wi-Fi and other untrusted networks, and our PTaaS platform keeps that testing continuous as your fleet and policies change.
Key Takeaways
- Since at least June 2026, attackers have compromised hotel Wi-Fi and venue gateways to poison DNS for every guest
- Researchers directly observed device code abuse that gave attackers MFA-satisfied Microsoft 365 tokens with no phishing email
- Adversary-in-the-middle theft and WPAD success are not confirmed for this campaign, and the APT28 link is an assessment only
- Public resolvers like 8.8.8.8 do not help; only full-tunnel VPN or strict-mode encrypted DNS stops the poisoning
- Block the device code flow, move travelers to FIDO2 or passkeys, and review Entra ID logs after trips
Frequently Asked Questions About the Hotel Wi-Fi Attack
How does the hotel Wi-Fi attack steal Microsoft 365 access?
Attackers take control of the captive portal gateway and poison its DNS answers, so a traveler’s device goes to attacker infrastructure instead of Microsoft. In the confirmed cases, a user who approved a device code prompt handed the attacker OAuth tokens.
Does MFA stop this attack?
Standard MFA does not stop device code abuse, because the user approves a real Microsoft prompt and the attacker receives MFA-satisfied tokens. Phishing-resistant FIDO2 keys and passkeys bind sign-in to the real domain and resist this class of attack.
Who is behind the campaign?
No one has been formally named. Researchers judge the tradecraft similar to APT28, a Russian military intelligence group, but found no shared infrastructure or code, so treat that as an assessment, not attribution.
Does using 8.8.8.8 or another public resolver help?
No. Researchers found that hard-coded resolvers did not stop the poisoning, because the queries traveled unencrypted through the compromised gateway. Only encrypted DNS in strict mode or a full-tunnel VPN prevents it.
Which control matters most?
Always-on, full-tunnel VPN with no split tunnel for DNS or authentication traffic. It sends every lookup to your own resolver before the hotel gateway can touch it.
What should security teams check after travel?
Review Microsoft Entra ID sign-in logs for device code sign-ins, new OAuth grants, unfamiliar IP addresses and sessions created during or after a trip, and search DNS and proxy logs for the published domains.

