In a report released last month on "next-generation endpoint security," Enterprise Security Group (ESG) analyst Jon Oltsik observed that enterprises and antivirus vendors alike find keeping up with endpoint security difficult because of the sophisticated nature, prolific volume, and exponentiating complexity of attacks.
"40% to 50% of new sophisticated malware attacks could evade endpoint AV, compromise PCs, and act as a beachhead for advanced cyber-attacks," wrote Oltsik. "CISOs realize today that, regardless of the controls they deploy, some malware will sneak through, so they need continuous monitoring and visibility of endpoint behavior."
"Continuous," however, has practical limits.
ESG found that the top endpoint-security challenge -- as indicated by approximately 25% of the 385 cybersecurity professionals surveyed for ESG's report -- was respondents' InfoSec teams taking too long dealing with too many security alerts, many of which are "false alarms." This suggests a -- ahem -- continuous trend.
Endpoint-security obstacle #1: insufficient automation
ESG and other pundits have concluded that proper endpoint-security management demands enhanced automation and machine-learning tools -- with the clearing out of security alerts being but one use case.
In an Industrial Internet Consortium (IIC) whitepaper on best practices in endpoint security, published in March, IIC emphasized automated protocols as a common denominator for both secure endpoint identities and secure attestations. Automation driven by public-key cryptography standards (PKCS), reported IIC, is critical to ensuring safety and certainty in the digital supply chain of certificates, firmware updates, etc. -- helping to keep at bay any malware that might otherwise slip through (sub)standard AV solutions.
Indeed, 17% of ESG study respondents identified their AV software as their top endpoint-security challenge, while 19% pointed to too many manual processes because of their lack of integrated endpoint-security automation. (It should be noted that respondents were allowed to select up to two responses.)
When it comes to malicious bots and the like, some contend that fighting AI with AI can be a losing battle. Consequently, a wholesale ban on non-whitelisted bots can help cut down on endpoint-security alerts and keep them manageable.
Nonetheless, IIC maintains that automated endpoint updates should be reliable without in-house whitelists or blacklists (typically manually input) as a matter of scalability.
"The number of attacks on industrial endpoints has grown rapidly in the last few years and has severe effects," wrote Steve Hanna, co-author of the whitepaper. "Unreliable equipment can cause safety problems, customer dissatisfaction, liability, and reduced profits."
Endpoint-security obstacle #2: legacy devices
Despite these admonitions against equipment unreliability, the authors make allowances for legacy endpoints. Still, they concede (1) that some of the most effective endpoint-security measures are embedded in hardware (typically not an option for legacy devices), and (2) that inadequately secure legacy endpoints must rely on network-security measures.
This latter point rather defeats the purpose of endpoint security. Why struggle against the most modern, most secure endpoints when there are more vulnerable legacy endpoints to be pwned?
To be fair, it is feasible to implement lower levels of trust across legacy endpoints -- but perhaps impractical. Legacy endpoints are maintained for a reason (usually involving cost). Presumably, therefore, these legacy endpoints still need levels of accessibility appropriate for more up-to-date endpoints -- despite the confidentiality and integrity risks.
Endpoint-security obstacle #3: poor security culture
This preference among enterprises -- even among InfoSec workers -- to mortgage endpoint security for agility's sake is further evident in ESG's findings:
- Respondents' second-biggest endpoint-security challenge (23%) was that regular re-imaging of infected endpoint devices creates more work for respondents' helpdesks and "imped[es] end-user productivity."
- 17% of respondents also complained that "imped[ed] end-user productivity" caused by endpoint-security agents slowing down endpoint processes was their organization's top endpoint-security challenge.
- 14%, meanwhile, said that their top endpoint-security woe was lacking the budget for "the right endpoint-security products".
ESG is not alone in making such findings. Verizon Wireless , for instance, recently reported that nearly one third of surveyed mobility professionals admitted that their organizations sacrificed mobile security in favor of business agility -- at significant risk. (See: Verizon Mobility Security Index Shows Enterprises Not Doing Enough.)
More depressingly, in a world where enterprises can be split into the hacked and the unaware of being hacked, a doubtlessly overly optimistic 10% of ESG respondents reported having no endpoint-security challenges whatsoever.
- Cisco Warns of Possible Smart Install Client Hacking
- IoT Regulation Could Save the Internet
- Researchers Bring AI to Endpoint Security
—Joe Stanganelli, principal of Beacon Hill Law, is a Boston-based attorney, corporate-communications and data-privacy consultant, writer, and speaker. Follow him on Twitter at @JoeStanganelli.