McKesson breach: How a vishing attack allegedly led to large-scale data theft
In August 2026, McKesson Corporation, a major American healthcare services and pharmaceutical distribution company, disclosed a cybersecurity incident after the ShinyHunters extortion group claimed responsibility for stealing 284 million records. The attackers reportedly used voice-based social engineering to target employees.
McKesson confirmed unauthorized access to third-party applications and data exfiltration but hasn’t publicly confirmed the full attack chain or the claimed 284 million records.
The McKesson data breach shows how a convincing phone call can put large amounts of sensitive data at risk.
Key highlights
- McKesson discovered the cyberattack on Aug 25, 2026, and disclosed it publicly on Aug 28.
- The company confirmed data exfiltration involving a subset of customers in its Oncology & Multispecialty and Medical-Surgical units.
- ShinyHunters claimed it used voice-based phishing to compromise employee accounts and access third-party applications, but McKesson hasn’t confirmed the attack chain.
- The incident is a reminder for businesses to strengthen identity verification, access controls, monitoring, and third-party application security to reduce data theft risks.
McKesson data breach: What happened?
A cyberattack hit McKesson’s Oncology & Multispecialty and Medical-Surgical business units, and the company disclosed it publicly on Aug 28. During the attack, someone accessed certain third-party applications without authorization and exfiltrated data.
On Sep 8, McKesson issued a further notice saying that personal and health information may have been affected, including:
- Names, dates of birth, patient IDs, and contact information
- Health insurance details, including Medicare and Medicaid identifiers
- Medical records, diagnoses, treatment information, and billing or claims data
- In some cases, Social Security numbers and financial information
McKesson hasn’t confirmed the full attack chain or how much information was accessed or exfiltrated, so the breach’s full impact is still unclear. ShinyHunters claimed 284 million records were stolen, but that figure hasn’t been verified. HIPAA Journal’s Sep 18 update noted that McKesson had not disclosed the number of affected individuals at the time.
What data did ShinyHunters claim to have stolen?
ShinyHunters claimed that it extracted about 1 terabyte of data, including 284 million records, from McKesson’s systems. The group told BleepingComputer that the figure represented raw records, not 284 million unique patients. The volume and the types of data the group says it took are also unconfirmed.
According to the group, the stolen information included:
- Personal identifiers: Names, home addresses, email addresses, phone numbers, and dates of birth
- Healthcare identifiers: Patient IDs, medical record numbers, and Medicaid-related information
- Sensitive medical details: Diagnoses, medications, allergies, appointments, and treatment-related information
- Financial and identity data: Social Security numbers and other potentially sensitive personal details
Have I Been Pwned gives a clearer, independent view of the exposure. It found about 6.4 million unique email addresses in the leaked dataset (also cited in HIPAA Journal’s Sep 18, 2026, update). According to HIPAA Journal, these addresses belonged to patients, staff, marketing campaign recipients, and healthcare provider contacts.
Neither figure shows how many patients were affected.
How did the McKesson cyberattack reportedly work?
ShinyHunters claimed that the attack began with voice-based social engineering (vishing) and progressed to McKesson’s cloud applications. The group’s account, reported by BleepingComputer, suggested that attackers impersonated IT or help desk personnel to trick employees into granting access.
The attackers allegedly compromised Okta single sign-on (SSO) accounts, then used that access to reach Salesforce and Snowflake environments and extract data. McKesson hasn’t verified this attack chain or the claimed ransom demand.
Here’s a step-by-step overview of the reported attack:
- Voice-based social engineering: Employees were allegedly targeted with vishing calls.
- Okta SSO account compromise: Attackers allegedly gained access to employee identities or sessions.
- Access to Salesforce and Snowflake: Attackers reportedly moved into business applications and data environments.
- Data extraction and extortion: The group claimed it stole data and demanded a ransom.
ShinyHunters also reportedly used the domain mckesson[.]claims, which was set up to look like a corporate help desk address. There’s no confirmation from McKesson that the domain was used.
Why are healthcare organizations attractive targets for data theft?
Between Oct 21, 2009, and Apr 30, 2026, the U.S. Department of Health and Human Services Office for Civil Rights breach portal recorded 7,670 large healthcare data breaches affecting 500 or more individuals.
Healthcare organizations hold extensive personal, medical, insurance, and financial information, which makes them attractive targets for cybercriminals looking to steal data and extort organizations.
Attackers may exploit trusted relationships between patients, providers, insurers, and healthcare vendors to make fraudulent communications appear legitimate.
Several incidents in 2026 show these risks:
- Medtronic: In April 2026, Medtronic confirmed unauthorized access to certain corporate IT systems. ShinyHunters claimed responsibility and said it stole millions of records, though Medtronic didn’t confirm the attacker or how the attack worked. According to a filing with the Oregon attorney general, about 3.8 million people were affected.
- AdaptHealth: In July 2026, AdaptHealth disclosed a social engineering attack in which a third-party contractor’s user session was compromised. The incident exposed patient information, and AdaptHealth later said that about 4.1 million people may have been affected.
Also read: Apollo data breach: How social engineering can put sensitive data at risk
What does the McKesson breach teach security teams?
Although the attack chain is unconfirmed, the reported use of voice-based social engineering suggests that security controls need to address human manipulation as well as technical vulnerabilities.
Security teams can strengthen their defenses with the following checklist.
- Verify help desk callers through a separate, trusted channel: Do not rely only on caller ID or information given during a call. Confirm identity using an established and verified contact method before resetting passwords or changing account access.
- Use phishing-resistant multi-factor authentication (MFA): Roll out passkeys or FIDO2 security keys for privileged accounts and remote access. These controls offer stronger protection against credential theft than passwords or SMS codes.
- Limit access across connected SaaS applications: Give employees access only to the information and systems they need to do their jobs. Regularly check SSO permissions and restrict access to sensitive customer and healthcare data.
- Monitor unusual account activity: Set up alerts for logins from unfamiliar locations, unexpected permission changes, unusual file downloads, and large data transfers. Make sure the security team reviews and investigates these alerts quickly.
How does Diopter fit into live-call social engineering defense?
Traditional security controls protect accounts and systems, but they may not fully address the manipulation that occurs during a live conversation, when an attacker impersonates an employee, executive, or IT support representative to gain trust and get access.
Diopter monitors voice and video calls in real time. It can flag suspicious voice or video signals, mismatches between who a person claims to be and how they behave, and unusual conversation patterns. That gives employees and security teams a way to spot fake interactions, especially during help desk requests.
Diopter works alongside existing security measures like multi-factor authentication, access controls, and security monitoring, and helps employees recognize possible social engineering attempts during calls.
Start with a simulated attack on your own team to see where your process breaks. Get a free walkthrough
McKesson breach: Key takeaways
The McKesson data breach shows how social engineering can lead to serious data security problems. For example, an attacker could trick an employee over the phone into handing over account access, then use it to reach sensitive information.
The exact details of the attack and the reported theft of 284 million records haven’t been confirmed, but organizations should still verify identities, use strong multi-factor authentication, limit access to sensitive information, and monitor systems for suspicious activity.
Organizations still need traditional security measures, but they also need protection during phone and video calls. Diopter adds that protection by helping detect suspicious voice and video interactions as they happen.
Keep reading: social engineering
FAQs
How did McKesson respond to the breach?
According to McKesson’s notice, the company activated its incident response protocols, launched an investigation, and brought in external cybersecurity experts. It took steps to prevent further unauthorized access and continued monitoring its systems. The company also said it would offer free credit monitoring, identity protection services, and a dedicated information line to affected people.
What should affected customers know?
McKesson’s notice says the exposed information may include personal, health, insurance, and financial details. Customers and patients should follow official updates, watch for suspicious messages or calls, and stay alert for possible identity theft.
Did the breach disrupt operations?
McKesson said its distribution centers remained operational and it continued to ship products across its network. The company also said it believed there was no ongoing unauthorized activity in its systems.
How can healthcare organizations protect employees from vishing attacks?
Staff should check callers’ identities through a separate, trusted channel before resetting passwords or changing access permissions. Strong MFA and activity monitoring can limit the damage if a call succeeds, and tools like Diopter can help detect suspicious calls. To check a recorded call, use our AI voice detector.
Monthly analysis of AI social engineering, voice fraud and deepfake attacks on enterprises.
One email a month. No spam, and we never share your address.