
A Fake IT Message Opened Twilio’s Systems—and Reached Beyond Its Customers
A 2026 guilty plea in a corporate-phishing conspiracy brings a criminal consequence to a campaign that compromised the communications provider. The downstream exposure included verification codes, rather than the encrypted messages themselves.
In April 2026, a British man pleaded guilty in California to a conspiracy that began by making a company’s employees believe their own technology staff needed them to log in. The consequences extended well beyond stolen employee passwords. The scheme reached corporate systems and, ultimately, individual cryptocurrency holdings.
Tyler Robert Buchanan pleaded guilty on April 17 to conspiracy to commit wire fraud and aggravated identity theft. The Justice Department’s account of his plea described attacks against at least a dozen companies and theft of at least $8 million in virtual currency from individuals. That figure covered the conspiracy; it was not a bill for damage to a single company.
The department described the victims by industry. KrebsOnSecurity’s reporting on the plea named Twilio among the technology companies breached in the 2022 campaign. Twilio’s own account supplies a closer view of how the attack worked—and how access inside one supplier could create exposure for people using another service.
The messages arrived on employees’ phones and impersonated Twilio’s administrators. They directed staff toward pages that appeared to be the company’s login system. Some employees entered their credentials. According to Twilio’s final incident update, published October 27, 2022, the attackers then used those credentials to enter internal administrative tools and applications containing customer information.
Twilio ultimately identified 209 affected customer accounts out of a customer base exceeding 270,000, as well as 93 affected Authy end users. The last unauthorized activity it observed was August 9. Its investigation found no evidence that attackers had obtained customers’ console passwords, authentication tokens or API keys. That limitation did not mean the information inside every internal support tool had been protected: access to those tools was the route the company had disclosed.
Signal was one downstream customer. It used Twilio to verify telephone numbers. In its account of the incident, Signal said approximately 1,900 users potentially had either a registration code exposed or their number revealed as associated with the service. The attacker explicitly searched for three numbers; one of those users reported that their account had been re-registered.
This was not a breach of Signal’s stored message history. Signal said previous messages, contacts and profile information remained protected. But the ability to register a number elsewhere could allow someone to send and receive new messages under it. Verification of who could use the account was a separate boundary from encryption of conversations already stored on a device.
Signal required the potentially affected users to register again and completed direct notifications by August 16, 2022. A compromise at its verification supplier had produced a response inside its own service, even though its central encryption protections had not been broken.
A related attempt against another American technology company produced a different outcome. Cloudflare’s August 2022 incident account described messages reaching at least 76 employees in less than a minute. Three employees entered credentials on the false login page. The company said physical security keys, required for its applications, prevented those credentials from becoming successful logins. It found no systems compromise.
The distinction was a control, rather than an absence of human error. Cloudflare described keys tied to the legitimate service, so an imitation page could not supply the additional authentication needed. It reset exposed credentials and sessions. Its account offers a comparison within a campaign of similar-looking attacks, without establishing that every company faced identical systems or circumstances.
Twilio later recorded its own control changes in a quarterly securities filing: stronger authentication requirements, additional controls within its virtual private network, reduced access to some internal tools and more frequent reauthorization. The filing described remediation following the 2022 breach, rather than a new intrusion.
The 2026 plea added an admitted criminal conspiracy to a record previously dominated by companies’ incident reports. Its consequences had traveled through several boundaries: an employee’s phone, a corporate administrative system, a customer’s verification service and an individual’s account. At each step, the decisive question was which authority the stolen credential could carry into the next system.
Compiled November 16, 2026. Records and findings keep the dates stated in their sources; each link in the text opens the document it relies on.