→ What CIS is
CIS is the Center for Internet Security. Its Critical Security Controls distil the most effective defensive actions into a single ordered list. The current release, v8.1, keeps the 18-control structure introduced in v8 and refines it — including better alignment to the NIST Cybersecurity Framework and updated safeguard descriptions.
The big shift in v8 was moving from device-centric thinking to activity-centric, and folding the old "sub-controls" into safeguards — there are 153 of them in total. The controls also map across to many established standards and frameworks, including the NIST CSF, NIST SP 800-53, the ISO 27000 series, PCI DSS and HIPAA.
→ The five functions
CIS aligns to the same five functions you'll recognise from the NIST Cybersecurity Framework — a clean way to think about where each control sits in the lifecycle:
→ Implementation Groups (IG1–IG3)
Not every organisation needs all 153 safeguards on day one. CIS sorts them into three Implementation Groups by maturity and risk, so you can start with the essentials and grow into the rest.
→ The 18 Controls
The full ordered list. The IG badge shows the lowest implementation group in which each control first appears.
→ CIS Benchmarks
Separate from the Controls, CIS Benchmarks are configuration baselines — consensus-built, best-practice hardening guides for specific technologies. There are 100+ benchmarks covering more than 14 technology groups.
| Group | Examples |
|---|---|
| Operating systems | Windows (desktop & server), Ubuntu, Debian, RHEL, SUSE, macOS, Solaris, AIX |
| Server software | IIS, NGINX, Apache, Tomcat, MS SQL, PostgreSQL, MongoDB, Oracle DB, Docker, Kubernetes |
| Cloud providers | AWS, Microsoft Azure, Google Cloud Platform |
| Mobile | Apple iOS, Android |
| Network devices | Cisco, Juniper, Palo Alto, Check Point Firewall |
| Desktop software | Chrome, Edge, Firefox, Safari |
Each benchmark recommendation references one or more CIS Controls, and each goes through two phases of consensus review before publication. They're the practical "how do I actually configure this box" layer beneath the Controls.
→ CIS-RAM — the risk model
CIS-RAM (Risk Assessment Method), currently v2.1, is the method for assessing how well your security posture implements the Controls against your acceptable level of risk. Its guiding principles are refreshingly plain:
- Risk analysis must consider the interests of all parties that could be harmed by the risk.
- Risks must be reduced to a level that would not require a remedy to any party.
- Safeguards must not be more burdensome than the risks they protect against.
That last principle is the useful one in practice: it gives you a defensible way to say "this control isn't proportional here" — the same logic that underpins risk treatment in ISO 27005.
→ How it fits with ISO and NIS2
CIS and ISO aren't rivals — they're complementary layers. ISO 27001 gives the governance and management-system frame; CIS gives the prioritised, technical "do this first" detail; CIS Benchmarks give the box-level configuration.
- CIS Controls map directly onto many ISO 27002 / Annex A controls — e.g. CIS 1 & 2 sit alongside Annex A asset management.
- For NIS2, the Controls are a fast, concrete way to satisfy the directive's risk-management measures without starting from a blank page.
- Drive control selection with risk via ISO 27005 or CIS-RAM, and govern it all under ISO 27001.