A breach is unauthorized access to systems, data, or networks that compromises confidentiality, integrity, or availability. It occurs when someone gains entry without permission, often exposing private or protected information.
Breaches are not accidental glitches. They result from deliberate actions-such as hacking, insider misuse, or software vulnerabilities-and trigger legal, financial, and reputational consequences. The term applies across sectors: healthcare, finance, government, and tech.
According to the U.S. National Institute of Standards and Technology (NIST), a breach involves "an event that results in the unauthorized disclosure, modification, or destruction of information." This definition excludes routine system errors or user mistakes unless those errors enable unauthorized access.
Most breaches follow a predictable sequence: reconnaissance, exploitation, persistence, and exfiltration. Attackers scan for weak points-like outdated software or reused passwords-then exploit them to gain entry.
Once inside, they move laterally through networks, escalate privileges, and extract data. Tools like malware, phishing emails, or stolen credentials enable this process. In 2023, 61% of breaches involved compromised credentials, per Verizon's Data Breach Investigations Report.
For example, in 2021, a hacker accessed a major U.S. hospital network using a single employee's password. The attacker remained undetected for 112 days, copying patient records, including Social Security numbers and medical histories. This was a breach-not a system crash-because it involved unauthorized data access.
Not all breaches are the same. They vary by method, target, and intent. The most common categories include:
| Type | Method | Example |
|---|---|---|
| Data theft | Exfiltration of personal or financial data | 2020 SolarWinds supply chain attack |
| Malware infection | Ransomware encryption of files | 2021 Colonial Pipeline ransomware incident |
| Insider threat | Employee misuse of access rights | 2019 Facebook data leak by contractor |
| Phishing | Fake login pages trick users into sharing credentials | 2022 Microsoft email phishing campaign |
Each type requires different detection and response strategies. For instance, a phishing breach may be stopped early by training employees to spot suspicious emails. A supply chain breach, like SolarWinds, demands continuous monitoring of third-party software updates.
Organizations store vast amounts of data-customer records, financial details, intellectual property. When security controls fail, attackers exploit gaps. Common causes include poor password policies, unpatched software, and insufficient access controls.
High-profile breaches often stem from overlooked risks. In 2020, a major retailer suffered a breach after an employee used a personal USB drive to transfer files. The device carried malware that spread across the network. This wasn't a sophisticated hack-it was a failure to enforce basic data handling rules.
Regulatory frameworks like HIPAA (healthcare) and GDPR (Europe) define breach thresholds. Under GDPR, organizations must report a breach within 72 hours of discovery. Failure to comply can lead to fines up to 4% of global annual revenue.
Many people confuse a breach with a system outage or a software bug. A server crash doesn't count as a breach unless someone exploited it to access data. Similarly, a misdelivered email containing sensitive information is a data mishandling issue-not a breach unless the recipient accessed it without authorization.
Another mistake is assuming that encryption prevents breaches. While encryption protects data at rest, it doesn't stop attacks that bypass authentication. In 2022, a cloud storage provider experienced a breach despite encrypting all stored files. Attackers gained access via a leaked API key, proving that encryption alone isn't enough.
There's also confusion between a breach and a vulnerability. A vulnerability is a flaw in software or configuration. A breach happens when someone exploits that flaw. Think of a vulnerability as a locked door with a broken lock. The breach occurs when someone walks in.
When a breach occurs, the immediate priority is containment. Organizations isolate affected systems, disable compromised accounts, and conduct forensic analysis to determine the scope.
I once observed a small nonprofit respond to a breach involving donor data. An employee clicked a phishing link, allowing attackers to access the database. Within 90 minutes, we disabled the account, changed all passwords, and notified affected individuals. We also reported the incident to the Federal Trade Commission under the FTC Act.
Afterward, we reviewed our incident response plan. The breach revealed gaps in staff training and multi-factor authentication. We implemented mandatory cybersecurity modules and required MFA for all external logins. These changes reduced future risk.
Under the law, organizations must notify affected parties when a breach exposes personal information. Notification timelines vary by jurisdiction. In California, businesses have 30 days to inform consumers after discovering a breach.
Not all data exposure is a breach. If a company sends a customer's name and address to a third party for shipping, that's data sharing-not a breach. But if the same data is sent to a malicious actor who uses it for identity theft, that's a breach.
Also, a breach doesn't require data loss. Simply gaining access to a system-without stealing or altering data-can qualify. For example, if a hacker logs into a corporate network and views internal documents but deletes nothing, it still counts as a breach under NIST guidelines.
Finally, a breach is not the same as a cyberattack. All breaches involve cyberattacks, but not all cyberattacks result in a breach. A denial-of-service attack disrupts service but doesn't necessarily expose data. It's a disruption, not a breach.
Understanding what is a breach helps organizations prioritize security measures. It's not about preventing every incident-it's about detecting and responding quickly to limit damage.