A consortium of US federal agencies have drawn up a list of critical security controls they hope will serve as a gold standard for cybersecurity.
The Consensus Audit Guidelines (CAG) list is part of larger plans to apply the CSIS Commission report on cybersecurity as a blueprint for making information security systems more secure. A public consultation on the scheme, launched on Monday, is due to run through 23 March. After that point federal security agencies will road-test the scheme.
Information security specialists at federal agencies pooled their knowledge on current attack techniques and countermeasures to draw up a list of 20 key actions, termed security controls, that organisations need to take to defend against assault. The first 15 controls on the draft list lend themselves to automation, while the remaining five have more to do with broader security policy and personnel issues.
Although these controls were drawn up by federal agencies they might be applied across diverse industry sectors from retailing, to banks, defense contractors and government agencies.
Each of the 20 controls was matched against specific threats and benchmarked in what might be considered a closed alpha test that its developers want to open up for wider scrutiny. It's hoped the controls, once fully vetted, will become the "standard baseline for measuring computer security". Most of the controls are not prescriptive and talk in broad terms about the need to audit equipment, use up to date ant-virus defences, firewalls and the like.
The list resembles the guideline drawn up by the credit card industry for adherence to the PCI DSS, at least at first sight. Experts involved in the scheme argue that CAG is far more ambitious.
"I do not know of anything going on in security that will have the impact this one can have," Alan Paller, director of research at the SANS Institute told El Reg. "It's a complete revolution in federal cybersecurity, and business security as well."
"In the past cybersecurity was driven by people who had no clue of how the attacks are carried out. They created an illusion of security. The CAG will turn that illusion to reality".
"If the nation (and the rest of the developed world) cannot make the CAG work we will continue to fall further behind the attackers at an accelerating rate," he added.
The CAG project is led by John Gilligan, former US AirForce chief information officer and began last year in response to data losses in the US defense industry. Organisations involved include the National Security Agency, the US Department of Homeland Security, US-CERT, the US Department of Defense, the US Department of Energy Los Alamos National Lab and three other National Labs, among others.
The Mitre Corporation, which runs the CVSS vulnerability scoring system, the SANS Institute and commercial penetration testing firms are also involved.
Proposed controls not already on the list need to be provably effective in either blocking or mitigating against known attacks.
The resulting guidelines will be evaluated against other security stands such as ISO 2700x, HIPAA (health care regulations), PCI, and SOX (Sarbannes-Oxley) to see whether actions included in these standards are better than approaches suggested in draft CAG guidelines. If so, they will be taken on board, as appropriate.
Dan Galik, Chief Information Security Officer, US Department of Health and Human Services, welcomed the proposals as an improvement on existing federal security regimes.
"I think it will go a long way towards recalibrating the Federal cyber security efforts away from being what many have described as a report card driven paper-work exercise, to instead being now properly focused on meaningful efforts to improve the real security posture of our operational systems," he said. ®
Critical security controls
- Hardware audit.
- Inventory of authorized and unauthorized Software.
- Secure configurations for computers and servers
- Secure configurations of network kits such as firewalls and routers.
- Boundary defense
- Maintenance of audit logs
- Application software security
- Application of administrative privileges
- Access controls based on need to know
- Continuous vulnerability testing and remediation
- Dormant account monitoring and control
- Anti-malware defenses
- Limitation and control of ports, protocols and services
- Wireless device control
- Data leak protection
- Secure network engineering
- Red team exercises
- Incident response capability
- Data recovery
- Security Skills Assessment and Training