As departments increasingly try parsing mammoth streams of Web activity to detect cyber threats, expect attackers to fight back by gaming the big data analytics, according to a new cybersecurity report by Georgia Tech researchers.
Even U.S. spies could be fooled.
"Government agencies, such as the National Security Agency, collect and process information to look for threats to national security," states the report, which was released Wednesday. "As the use of such big data analytics spreads, attackers will have to find ways to hide from statistical analysis and anomaly detection."
The researchers project that agencies will face the problem of big data manipulation in the next five to 10 years.
An attacker could program fake malicious operations to exhaust the energy of security analysts, or slowly create a wide variance in some metrics to make it look like odd network behavior is normal, they hypothesized.
"Defending against such attacks requires that defenders be able to detect very slow changes in the data and be able to flag such changes as suspicious," the report states.
To hear more tips, join us at Nextgov Prime in Washington on Nov. 20-21, where I'll be discussing some of the trends we're seeing in data breaches and the implications for federal networks. Registration is free for federal employees.