Second-factor codes sat in reach after someone walked into Apify with a compromised credential. Passwords, the company told customers, remained safe. API tokens and proxy passwords did not get that luxury. According to BleepingComputer, an unauthorized third party used a compromised credential on 9 and 10 November 2026 to reach parts of Apify’s internal systems, obtain additional account credentials, and access related Apify accounts. Apify detected the activity at 12:13 UTC on 11 November 2026, cut off access, rotated customer API tokens and Apify Proxy passwords, reset active sessions, and advised affected customers to reset 2FA because codes may have been exposed.
Public reporting still does not establish how that first credential was compromised. Containment that followed addressed how far a transferable internal login had already reached once it worked.
What the compromised credential unlocked inside Apify
Apify’s own incident language is direct. According to the company’s update as reported by BleepingComputer, on 9 and 10 November 2026 an unauthorized third party gained access to parts of Apify’s internal systems. From that foothold the attacker obtained additional account credentials and reached related Apify accounts. The Apify status page and customer notices describe the same sequence: internal access first, then broader account reach, then containment.
Apify serves roughly 80,000 customers. Named logos in public coverage include Microsoft, Samsung, T-Mobile, Accenture, Siemens, the European Commission, and Roche. That customer base is the population that could be affected when internal credentials and platform account material sit in the same operational plane.
Detection landed on 11 November 2026 at 12:13 UTC. Apify invalidated internal credentials, rotated customer API tokens, rotated Apify Proxy passwords, and reset active sessions. Some third-party OAuth connections, including GitHub, were revoked as a precaution. Investigation found OAuth was not the attack vector. No further attacker activity was reported after access was cut off.
The initial compromise method is not in the public record. Public reporting has not established a proven path such as AiTM phishing, a spoofed login page, helpdesk recovery abuse, password spraying, or credential stuffing. What is established is that a credential worked on an internal path, then more credentials were obtained from that position.
| Control surface | What happened | Why it failed here |
|---|---|---|
| Internal login credential | Used by unauthorized party on 9 and 10 Nov | Transferable secret still worked |
| Additional account credentials | Obtained from internal foothold | Reachable secrets after first login |
| Customer 2FA codes | May have been exposed | Recoverable second-factor material |
| API tokens and proxy passwords | Rotated after detection | Long-lived secrets after foothold |
| OAuth (e.g. GitHub) | Revoked as precaution | Not the attack vector |
Why customers reset 2FA while passwords stayed safe
Apify’s customer letter, as reported by BleepingComputer, told affected users that 2FA codes may have been exposed while assuring them that passwords remained safe. That distinction is what the company communicated operationally. Assurances that passwords remained safe do not address second-factor codes that may have been read, copied, or replayed once an attacker was already inside systems that hold account recovery material.
Customers were told to reset 2FA. Individual notifications continued after the 12 November 2026 public reporting wave. One customer notice said there was no indication that account data stored on the platform had been affected. Which customer accounts were involved and what stored data was implicated remained under investigation. Quantified counts of hit accounts are not in the public sources used here.
Second-factor codes, seeds, or recoverable secrets create an additional exposure surface beside the password. When those codes can appear in internal systems an attacker already reached, password safety alone does not clear the second-factor risk. The login path still depended on transferable material. Public reporting does not name the exact MFA product Apify used for every account path. It does show that 2FA codes were treated as potentially exposed artifacts.
API tokens, sessions, and the post-detection scramble
Apify’s containment work after 11 November was classic cloud SaaS incident response. Customer API tokens were rotated. Apify Proxy passwords were rotated. Active sessions were reset. Selected OAuth connections were pulled down even though OAuth was not how the attacker got in. That last step was precautionary hygiene; investigation found OAuth was not how access was gained.
API tokens and live sessions sit after a successful authentication. Once the attacker held internal position and related account credentials, those long-lived secrets and sessions became the practical keys to customer-facing automation. Cutting them off stops reuse. It does not rewind the days the credential path was live. Shortening token lifetime after the fact would not have blocked the original internal login.
Workforce identity failures cascade into platform customer impact when internal credentials, account databases, and API control planes share trust. Apify’s customers run scrapers, actors, and integrations that hang off those tokens. A compromised internal credential is therefore not only an employee-account story. It is a customer-token story the moment the foothold can read or mint the secrets those customers depend on.
Two separate problems apply. Stopping the next transferable internal credential from working is an authentication design problem. Revoking API tokens, proxy passwords, sessions, and precautionary OAuth grants is containment after the login already succeeded. Both matter. Only the first prevents a window like 9 and 10 November from opening.
Related reading on preventing this class of internal credential and 2FA-code exposure is available on mfa2point0.com: how MFA 2.0 would change Apify-style credential and 2FA exposure.
FAQ
How did attackers get into Apify in November 2026?
Attackers got into Apify when an unauthorized third party used a compromised credential on 9 and 10 November 2026 to access parts of Apify’s internal systems. According to BleepingComputer’s reporting of Apify’s update, that foothold was then used to obtain additional account credentials and access related Apify accounts. Public sources do not establish how the first credential was originally compromised.
Were Apify customer passwords stolen in the breach?
Apify told customers their passwords remained safe while warning that 2FA codes may have been exposed. According to the customer letter summarized by BleepingComputer, affected users were advised to reset 2FA, not because passwords were confirmed stolen, but because second-factor codes may have been reachable after the internal intrusion.
Why did Apify rotate API tokens and Apify Proxy passwords?
Apify rotated customer API tokens and Apify Proxy passwords after detecting the intrusion on 11 November 2026 and cutting off attacker access. Those secrets and live sessions could be reused once the attacker held internal position and related account credentials, so rotation and session reset were required containment even though the initial path was a compromised credential rather than a direct API exploit.
Was OAuth or GitHub the attack vector against Apify?
OAuth was not the attack vector in the Apify breach. Apify revoked some third-party OAuth connections, including GitHub, as a precaution after detection. Investigation found OAuth was not how the unauthorized party gained access.
How many Apify customer accounts were affected?
Public reporting has not published a quantified count of affected customer accounts. Apify has roughly 80,000 customers, individual notifications were ongoing after 12 November 2026, and which accounts were involved and what stored data was implicated remained under investigation, with at least one notice stating no indication that platform-stored account data had been affected.