Former Homeland Security Secretary Michael Chertoff sees today’s security threats as one connected landscape. Speaking with Margaret Hu, a professor at William & Mary Law School, at the American Bar Association’s annual national security law conference in Washington, he described the domains of security as “all linked in some way or another.”
Foreign national groups, domestic terrorism, domestic crime, economic security and supply chains all affect national security, even though each may have distinct roles and practices. “We need to take a more comprehensive strategic view of our security challenges,” Chertoff said.
Technology has accelerated that convergence. Many nation-state adversaries back terrorist groups, and transnational criminal groups work in conjunction with them. “One thing the technology has done is it has more widely distributed the ability to cause economic harm and harm to human beings that used to be reserved for nation-state adversaries,” he said.
A comprehensive strategy and the cost of silos
Chertoff led the US Department of Homeland Security from 2005 to 2009, as the young agency’s components unified under one mission. He still sees an integrated approach as essential.
“Having an integrated, comprehensive strategy for dealing with security is essential,” he said. “You cannot treat each security issue as a standalone, here today, gone tomorrow.”
Distinct threat groups cross-pollinate and learn from one another, he said. The techniques used to uncover and neutralize those threats apply across the board.
“The biggest challenge you have with national security is the tendency of everybody to look at their own little silo and not consider the way things operate across the entire landscape,” Chertoff said. A strategic view of that landscape, he added, is “a necessary predicate before you drill down into the particularities.”
AI and the individual
AI can now learn what a person is interested in, what they will respond to and what will anger or upset them.
“The amount of data that can be absorbed and acted upon by AI far exceeds what any human being can even in their wildest dreams imagine doing,” Chertoff said. Before AI, using a database on a million people for nefarious purposes would have required thousands of people to digest it. Today, AI can absorb all of that data, integrate it and make judgments or decisions based on it.
What people do online, where they travel and the devices they rely on all generate data about them.
“What AI has allowed to happen now is the ability to take all of the data you generate … and make it into a tool that can be used to manipulate you and only you,” he said. “You can get messages on your phone or news on your phone that no one else will see.”
Fighting bad speech with good speech
Chertoff noted the rise of disinformation at scale, much of it aided by AI. He drew a firm line on how government should respond.
“I would be very nervous about having the government regulate disinformation in terms of content,” he said. Government can instead disclose facts that shed light on a message’s veracity and provenance.
“What you should be able to do is say this is not real. This is something that was generated by artificial intelligence,” Chertoff said. “It is fighting bad speech with good speech, which has always been a principle of the Constitution.”
Civil liberties in the age of AI
“We are facing some really significant challenges in the area of the balance of civil liberties and security,” Chertoff said. Commercial incentives add to that pressure, which he described as “people’s desire to make money.”
He called AI’s capacity for individual manipulation “a major risk to civil liberties.” He also pointed to concern that China could occupy critical nodes on US networks, collect information on Americans and use it to promote its agenda.
Chertoff favors general government standards for AI. He cautioned that micromanaging development could stifle innovation and cede ground to China in both the commercial and national security spaces.
Optimism about the American public
Chertoff acknowledged worries about technology getting away from us and about people feeling powerless. He also sees reason for confidence.
“If you look at the history of human beings, we have had backsliding, but in general, there is sometimes a pushback against the backsliding and people wind up recovering their values,” he said. “I have generally an optimistic view of the American public, that I believe will push back against the sense that people have that they are powerless.”
Implications for organizations
At The Chertoff Group, we see direct implications for companies. Customer and employee data is a target for adversaries with AI capabilities. Network infrastructure and vendor access deserve the same scrutiny as any other critical asset. Data governance is a security function and belongs in every risk strategy.
To learn more about what organizations must do to defend themselves in the age of AI, read our latest brief, Defending at Machine Speed.
Watch the full interview here.
Michael Chertoff is co-founder and executive chairman of The Chertoff Group.
