Monday Brief: 01 March 2021
A big tech week on Capitol Hill; GCHQ enlists AI; and space is being disrupted
What’s New: Last week, Congress held three important hearings on US national security and technology. Here are some highlights and thoughts on each.
Why This Matters: The 117th United States Congress is beginning its work, and emerging tech and “the cybers” are getting a lot of early attention.
Hearing #1: House Armed Services Committee — Defense Innovation
As I mentioned last week, I joined the Hon. Christine Fox — former Acting Deputy Secretary of Defense — and Dr. Victoria Coleman — Former DARPA Director — in testifying before the HASC on Defense Innovation.
What I’m Thinking:
We’re on the same page. There was broad agreement on the committee and among the witnesses that the challenge of integrating defense innovation isn’t really about acquisition authorities — it’s about culture and countervailing incentives that favor cost accounting, regulatory compliance, and administrative ease.
We have this culture for a reason. The DoD is arguably the most forward-looking element of the US government. It is also the most heavily engaged and its missions are often of the highest importance. This creates a “tyranny of the urgent” dynamic, where we have pockets of excellence in planning for the future but they are ensconced within an organization that has an unending list of requirements in the present — syphoning attention and resources.
Evolve or die. The prioritization of present requirements is understandable and largely inescapable (and our challengers have this problem too). But we are moving into a period where first mover advantage in some technologies could be decisive, so we have to walk and play chess at the same time.
Hearing #2: Senate Armed Services Committee — Emerging Tech and Their Impact on National Security
Witnesses: Eric Schmidt (former Google CEO), Brad Smith (President, Microsoft), and General Herbert J. Carlisle, USAF (Ret.; President of the National Defense Industrial Association).
Video and submitted testimonies can be found here.
Schmidt, who is on the Pentagon’s Defense Innovation Board and leads the National Security Commission on AI, made several important points:
On AI: “In the United States, we believe we are one or two years ahead of China, not five or 10 … Because of the diffusion of the technology, you have to expect that anything that’s invented in the open source AI world will immediately be adopted by China.”
On facial recognition: “Their technology is generations ahead of what is possible in the West … The threat is very, very real.”
“The [US] government will need to help with some forms of funding, and we need to let the private sector build those things and make it successful.”
What I’m Thinking:
Nuance matters. I agree with most of what Schmidt and others said about the nature and the challenge of China’s technological rise. But, it’s always good to season these assessments with the salt of nuance. Specifically, AI innovation isn’t monolithic and explaining who’s “winning” and who’s “losing” isn’t always clear. AI capacity and leadership largely comes down to three aspects: hardware (e.g. AI specialized semiconductors), data availability, and advancing algorithms.2 At different points within each of these aspects the US and China assert different advantages and disadvantages. Generally speaking, the US holds the advantage; but, Beijing’s advantages are accumulating and American leadership cannot be assumed.
Some things, however, are plainly true. Our relative technological edge is shrinking and appears to be doing so more quickly. The American system may be more dynamic and produce more novel AI discoveries, but China is breaking records when it comes to integrating these advancements and applying them to government and economic priorities. Critically, it’s important to remember that you don’t have to invent a technology to be able to use it more effectively.
Hearing #3: Senate Select Committee on Intelligence — Hacks of US Networks by a Foreign Adversary
Witnesses: Kevin Mandia (CEO, FireEye), Sudhakar Ramakrishna (CEO, SolarWinds), Brad Smith (President, Microsoft), and George Kurtz (President/CEO, CrowdStrike).
Testimony and video can be found here.
Not a lot new here, but still an important hearing.
Money Quote: “The attacker is the only one who knows the entirety of what they did.” — Brad Smith
What I’m Thinking: Nearly three months in and we’re still only at the beginning of this story. We don’t know what other compromises are associated with the “Holiday Bear” op, which of these compromises are still active, or what was actually done once a target was popped. Some of you know the technical details better than I do (heck, some of you are part of the investigation of this operation); but, if you’d like to have a better grasp of hacking basics, check out this explainer I coauthored. Or, here’s the TL;DR3 of what’s called “the cyber kill chain”:
Developed by Lockheed Martin, the cyber kill chain provides a framework for thinking about the various stages of a hack. The stages themselves are reconnaissance, weaponization, delivery, exploitation, installation, command and control, and actions on objectives. Depending on the sophistication of the hacker or hacking organization, and the relative strength of the defender, some of these phases may be longer than others. For example, an advanced hacking group may spend months in the reconnaissance phase looking for a vulnerability, while a criminal may happen upon a vulnerability through a phishing e-mail that someone clicks.
British Spies Enlisting AI
What’s New: GCHQ Director Jeremy Fleming has written an op-ed for the Financial Times (paywall), saying AI is critical for “allowing analysts to deal with ever-increasing volumes of complexity of data” and “improving the quality and speed of decision-making.”
Why This Matters: The article is a bit of public relations, reassuring the word that the agency’s use of AI will be “legal, proportionate, and ethical.” But it also seems to legitimately herald a new era for AI-enabled espionage for UK.
The article is part of a roll-out of the British spy agency’s new report, The Ethics of AI.
AI will be augmenting key missions including cybersecurity, spotting state-backed information operations, and tracking criminal syndicates around the world.
The public announcement, according to the Financial Times, reflects “growing anxiety that adversaries such as Russia and China are already weaponising AI technology against Britain and its allies.”
“In the hands of an adversary with little respect for human rights, such a powerful technology could be used for oppression,” wrote Fleming. “Inaction can let those who build the technology of tomorrow — whether a country or company — project their values or interests by stealth, poor design or inadequate diversity. The consequences are hard to overstate.”
What I’m Thinking: AI is getting more powerful and the scale and speed of modern threats will require its use in virtually every aspect of national defense. But this evolution brings its own challenges. I discussed these challenges in a 2018 speech you can watch below. (Which reminds me, I totally need to write this up.)
Bottom lines IRT AI-enabled intelligence and warfare: (1) We’re going to know more than we’ve every known; (2) We won’t always know how we know things; (3) AI will be authoritative … until it isn’t; and (4) the scale/speed of modern threats will necessitate automation.
Space Is Being Disrupted
What’s New: Elon Musk’s SpaceX is preparing a new space mission manned entirely by civilians.
Why This Matters: This is the latest demonstration of how government has lost its monopoly on space.
Four individuals are planned to launch into space later this year and to spend several days orbiting the Earth.
NASA is just watching. The agency will not choose the mission members. It will not train them. And it does not own or operate the rocket.
One astronaut is a billionaire who is funding the mission; another a physician assistant. The third will be randomly selected, and the last seat will go to the winner of a competition.
The Washington Post reports, “The commercial space industry is taking on ever more roles and responsibilities — flying not just cargo and supplies to the International Space Station, but even NASA’s astronauts there. The private sector will launch some of the major components of the space station NASA wants to build in orbit around the moon, and private companies are developing the spacecraft that will fly astronauts to and from the lunar surface.”
What I’m Thinking: Elon Musk, Richard Branson, Jeff Bezos, and other bajillionaires all see space as the new (profitable) frontier. And they’re right. Innovators, investors, and interested individuals are pursuing everything from overhead imagery to space tourism to asteroid mining and we’re just getting started. I suspect this will lead to dynamics that are similar to those we’re experiencing here on earth — government and private sector interests and capabilities will continue to intermingle, ultimately requiring private sector actors to choose a flag in order to protect those interests and capabilities.
That’s it for this Monday Brief. Thanks for reading, and if you think someone else would like this week’s newsletter, please share it with your friends and followers.
Have a great week!
Slang: Hotwash [Hot Wash]: An immediate "after-action" discussion and evaluation of performance and insights following an exercise, training session, or major event.
Andrew Imbrie, Elsa Kania, and Lorand Laskai, The Question of Comparative Advantage in Artificial Intelligence: Enduring Strengths and Emerging Challenges for the United States, January 2020, Center for Security and Emerging Technologies, accessed on February 26, 2021 at https://cset.georgetown.edu/research/the-question-of-comparative-advantage-in-artificial-intelligence-enduring-strengths-and-emerging-challenges-for-the-united-states/
Slang: "Too long; didn't read.", meaning a post, article, or anything with words was too long, and whoever used the phrase didn't read it for that reason.