Traditional spies under threat as CIA replaces them with AI

Image: MidCentArc - Flickr
Ingrid Fadelli

Spooks accelerate plan to stay ahead of the game with the widespread deployment of AI

The intelligence services have always been early-adopters. It should come as no great surprise then that, as Artificial Intelligence (AI) and robotics go mainstream, even human spies could soon be replaced or assisted by technological tools.

At a recent intelligence conference in Florida, the Central Intelligence Agency (CIA) has announced its plans to integrate the latest technologies into its daily operations.

For years, experts have been saying that Artificial Intelligence could soon become widely used among international intelligence agencies, making their operations faster, more effective, and harder to trace.

The recent advent of machine learning algorithms has opened countless possibilities for government agencies, who can now choose from a broad variety of surveillance tools and tracking platforms.

Tasks carried out by human spies, such as analyzing extensive video footage collected by CCTV, drones, or other feeds, could be taken on board by AI and machine learning algorithms, which have often been found to achieve similar results in far shorter time.

At a conference in San Antonio that took place almost a year ago, Robert Cardillo, director of the National Geospatial-Intelligence Agency (NGA), announced his intent to push for greater automation within the government organization.

The CIA’s deputy director for technology development, Dawn Meyerriecks, has announced her agency’s similar plans at this week’s 2018 GEOINT Symposium in Tampa, Florida.

During the conference, the CIA revealed that it is currently adapting to a new intelligence landscape in which the most threatening enemies are no longer human agents, but intelligent machines.

AI to evade eyes in the sky

After her keynote speech at 2018 GEOINT, Meyerriecks spoke to CNN about the new challenges posed by advanced technology and the ways in which the CIA is working to address these.

Until not long ago, CIA agents working overseas had to look out for foreign spies following them and trying to interfere with their missions.

But machine learning tools such as face and voice recognition are making hunting down adversarial officers much easier, without requiring humans to risk their lives and take physical part in such operations.

According to Meyerriecks, digital surveillance has now become remarkably advanced in almost 30 countries worldwide, turning physical tracking into an outdated procedure.

“Singapore has been doing it for years,” she told the CNN, adding that the CIA has been pursuing almost 140 AI projects over the past six months.

For one of the agency’s projects, a small team used a database of unclassified overhead and street view images to train machine learning algorithms, creating a map of cameras in a particular capital city.

This could help human agents to elude cameras, making them aware of specific places where they are under digital scrutiny.

Meyerrieck stressed the importance of using technology to evade digital surveillance, particularly at a time in which Russia is expelling American diplomats, and several spies or sources in China have been killed or are currently missing.

New ways of tracking sources or spies

A further challenge posed by new technology is the possibility of tracking people by analyzing their use of social media, smart phones, or other digital traces.

Earlier this year, for instance, a fitness tracking app called Strava was found to pose risks for security forces, as it could reveal their location or that of classified sites.

Other social media platforms, apps, or online activity could be equally useful for spy agencies trying to locate a particular place or person.

“Even if you turn your phone off 10 minutes before you get to your place of employment, do you think anyone’s fooled by where you’re going?” Meyerriecks told CNN.

In order to counteract these digital tracking techniques, the CIA is using increasingly advanced tools that send out incorrect information about agents’ location during the day.

The automated future of global intelligence

The CIA’s recent announcement at the GEOINT Symposium should not come as a surprise; America’s oldest spy agency has spent over 30 years preparing for a transition from human to computer agents.

A document dated back in 1984 talks about an “AI Steering Group”, specifically created to provide CIA executives with monthly reports about the state of AI research and development.

Now that AI is no longer an experimental or theoretical concept, it makes sense for the agency to start incorporating it into its operations.

While posing many new challenges for opposing agencies, AI can also result in faster and more thorough analyses, bringing global intelligence to a whole other level.

Several other countries worldwide are also embracing these profound changes, including China, Russia, and Singapore.

Russian President, Vladimir Putin, has often spoken about the huge impact that AI could have in the future. In an address to Russian citizens in late 2017, he said: “Artificial Intelligence is the future, not only of Russia, but of all of mankind. Whoever becomes the leader in this sphere will become the ruler of the world.”

Meanwhile, Russia has approved plan that aims to have 30% of automated or remote-controlled combat power by 2030 and has also started using new technology to boost its espionage and propaganda.

As the pace hots up, AI-fueled surveillance tools and other machine learning algorithms are likely to become a key part of the intelligence scene, leading to far more sophisticated tactics for international espionage.

Start typing and press Enter to search