HARTFORD, CT – Amid the debate over how much the state should regulate artificial intelligence in the private sector, state government is being protected from cyber attacks by AI technology that is making security decisions on the spot.
Cyber security companies CrowdStrike and Abnormal Security are two of nine firms used by state agencies that provide AI services. For some, such as Microsoft and Zoom, those services are a fraction of what they supply the state.
CrowdStrike and Abnormal Security have large roles in protecting state agencies against crippling attacks. The state’s annual contract with CrowdStrike is $1.3 million for the entire executive branch, or 31,200 licenses, according to Leigh Appleby, communications director for Connecticut Administrative Services.
MORE NEWS ON AI
CrowdStrike “uses AI and machine learning to provide real-time detection and response to a wide range of cyber threats, such as visualizing potential attacks in real-time.
Artificial intelligence is protecting state government from cyber attacks,” according to the state inventory of AI-related systems.
“It is incredibly powerful to continually evolve to protect and defend state technology,” Appleby said.
A state law went into effect last year that requires the Department of Administrative Services to conduct an annual inventory of artificial intelligence systems used by all state agencies. The inventory includes the agencies, the company providing the AI service, the uses of AI and details on decisions that the system can make, often without human intervention.
The Abnormal Security contract is for $372,363 and covers all executive branch email accounts, according to Appleby.
“Its behavioral analytics engine analyzes every email that comes into an organization, building an understanding of normal communication patterns,” the inventory says. “It then uses this baseline of normal activity to identify anomalies that may indicate sophisticated, socially engineered attacks or account compromise.”
“These companies are adding AI somewhat regularly to improve their products and reduce customers’ manual efforts,” Appleby said.
The use of AI is increasingly under scrutiny and its potential regulation is at issue as the legislative session grinds toward a conclusion.
The Senate passed a sweeping regulatory bill last week that has met criticism from Gov. Ned Lamont and Senate Republicans, who have said they are concerned, among other things, about putting additional burdens on small businesses. Lamont has expressed concerns that some provisions could stifle innovation and put the state at a disadvantage when trying to attract technology companies.
Senate Democrats say a law is needed to protect against discrimination in areas such as hiring and lending from AI algorithms and against things like so-called “deep-fake” videos that make it appear someone is doing something they are not.
Lamont’s office declined to comment Friday on the bill’s passage, referring to a previous statement in which it said Lamont remains concerned that this is a fast moving space and that we need to make sure we do this right and don’t stymie innovation.”
