ARTIFICIAL INTELLIGENCE: IMPLICATIONS AND CHALLENGES (FOR DEVELOPING COUNTRIES)

Parliamentary Intelligence-Security Forum
3 followers
Mar 18, 2026
28th Parliamentary Intelligence-Security Forum
Philippines🇵🇭- February 3-4, 2025
-Prof. Jamil Jaffer
Founder and Executive Director, National Security Institute (Antonin Scalia Law School in George Mason University) Board Member, U.S. Strategic Metals
AI Generated·May contain errors
Mr. Speaker, Congressman Pittenger, distinguished guests, thank you for being here today.
I'm here to talk to you about the advent of Generative Artificial Intelligence,
its opportunities, its implications, its challenges for our nations and for developing
countries around the globe. As we talk about the advent of Generative Artificial Intelligence
and all the things that come along with it, I think it's worth talking about how far we are
into this process. It was only in November 2022, just over two years ago, that ChatGPT 3.5
hit the public stage. We're talking about less than 24 months since this massive change that
we've seen take place around the globe when it comes to Artificial Intelligence. Since then,
we've seen the launch of LLM after LLM, Claude from Anthropic, Gemini from Google, Mistral in France,
Llama from Meta, Grok from X, Command-R from Cohere using Retrieval Augmented Generation,
and of course, just in the last two weeks, DeepSeq R1, an open source model from China
that has been all the rage in the last two weeks. Reinforcement learning and Retrieval
Augmented Generation have been a key part of the evolution of the latest version of LLMs,
as has test time compute and reasoning allowing models to use less time, less compute,
during the training phase and instead implementing it during the inference phase.
So today, I'm going to talk a little bit about some of these implications,
risks, challenges, and opportunities. But first, I want to talk about why you should even trust me.
And first, let me tell you first about why perhaps you shouldn't trust me. I have to admit,
I'm a lawyer, and perhaps even worse, I'm a lawyer that worked in U.S. national security for many
years, at the White House, the Justice Department, other places. So I've been trained to make
arguments that I know benefit my own cause and to tell you a lot or not tell you a lot about the
stuff that I know. At the same time, I'm also a recovering lawyer, which means that I'm apologizing
for all the bad things I did when I was a lawyer. And so maybe you can trust me a little bit more
than the typical lawyer. I'm also a venture capitalist, which means I put money to work
in this space, whether it's in cybersecurity, AI, or the like. So I'm oriented towards making
money for my investors, but our firm is also invested in doing good, supporting national
security, the United States, our allies, and our partners, including those here in the Philippines.
So maybe that cuts both ways as well. I'm an academic. I teach classes in this space, but
you know, I'm not a wonky academic. I try to apply a little bit of practice. So, you know, who knows?
And then finally, I'm a computer nerd at heart. My first computer was 1984, a TRS-80 color computer
with 4k of RAM, 4 kilobytes, mind you. We upgraded to 16 kilobytes, and it was the talk of the town.
So I love blinking lights. I love AI. I love this stuff. I'm a big nerd about it. So maybe that
means you can trust me there too. But let's talk about it, right? Let's jump right in. Let me say
right at the outset, I'm a believer and hugely bullish about the promise of AI for the globe.
Not just developed countries, but developing countries as well. This is because I believe
that AI and the increased pace and scale of adoption of AI will allow humans to continue
to specialize and get particularly good at the fine, deeply creative and novel thinking and
problem-solving that while AI will take much of the mundane and routine off the table.
I also want to say that I'm not a big believer in the claims made by some, including some of
the biggest and most prominent developers of AI and the ethics of its use, that
it is a problem that needs to be treated on the level of priority with other societal scale risks
like pandemics and nuclear war. I don't buy that. Let me say it in clear terms. I think those who
see AI as a potential global threat to humanity, our society, our civilization, or our planet,
the so-called AI doomers, they're just wrong. It doesn't mean I'm sanguine to the challenges that
AI presents. The real truth is there are huge issues that come along with AI. Job displacement,
misinformation, disinformation, inaccurate information, manipulation, deep fakes, privacy
and security issues, potential surveillance, the collection of biometric data, racial and
other biases including discrimination that goes along with those biases, lack of transparency
and explainability in AI systems, and of course the huge ethical issues that come with decision
making by AI or AI augmented systems, including in the space of weapons deployment and use in
the defense context. And these are just a few of those concerns associated with AI revolution.
So the question then as parliamentarians you might ask yourself is what if anything can or
should we do about these challenges and how can we take advantage of the opportunities that AI
presents without getting tangled up in all the problems? These are of course important questions.
After all, as I mentioned earlier, I think AI has a tide that can lift all boats.
In fact, I think that AI will have a disparate impact on populations primarily in developed
countries, particularly on jobs like lawyers, where it will replace a lot of the low-level
work that lawyers do and elite white-collar workers do. I think there are tremendous
opportunities for the developing world when it comes to AI. And in many ways, just as the
internet put the world's knowledge at your fingertips, at least for those people who
could access it, I think AI will make that information easier, more practical to use,
and provide more opportunities for the world's information to be distributed at scale.
But that still leaves open the question of how nations, including developing nations,
might capitalize on the benefits of AI without getting caught up in the challenges.
And let me answer that in two ways. Let me point to the parliamentarians who have felt so strongly
about these concerns that have acted rapidly to step in, fill the gap, and close the door to some,
or at least as they see it, all these threats. Now, of course, this is my friends from Europe,
and I hate to do it, but I'm going to use my friends from Europe as a foil. You are the bad
men and women of this story, because when it comes to AI, the European Parliament has done
what it always does. It's acted quickly, it's acted swiftly, it's regulated now and asked
questions later. In the barely two years since GPD broke upon the scene, the European Parliament
has passed a bill with 114 pages, 180 recitals, and 113 articles. Now look at the United States.
Congress has talked a lot about taking action, but hasn't acted yet. To be sure,
it's taken a lot of time to get up to speed, it's educated itself, it's held hearings and
discussions with the major AI companies, but it hasn't acted. To be fair also, the White House
did take some action. President Joe Biden came in with a 100-page EO, the largest one in history,
executive order, but that executive order, with some exceptions, largely didn't regulate,
had some reporting requirements, but by and large, the president's EO dealt with how the
government uses AI. And of course, President Trump, in the last week or two, has eliminated
that EO and called for a six-month review to assess what should be done. But this difference,
regulate on one hand versus educate on the other, legislate versus innovate, is a common trend
between Europe and the United States. And if you wonder what the results are, it's pretty
straightforward. The vast majority of technological innovations in the last two decades and all the
top 10 technology companies of the globe are headquartered in the United States. Nearly every
major AI LLM, with the notable exception of DeepSeek R1, and of course Mistral, is built by an
American company. And the deepest penetration of the use of LLMs around the globe, there's some
exceptions, some say China may have a deeper penetration, is in the United States, 20%
of the world's queries to chat GPT come from the United States. So what is the lesson for
the parliamentarians here? And I'll wrap up here. Perhaps take a beat, see how this plays out,
let's see how the innovation evolves. But as you think about, well these are real concerns,
these are not illegitimate concerns, I don't pretend to suggest they are. What might I do now?
Before, if I'm not going to act legislatively, what might I do as a legislator? Well one,
vote with your feet. As your government uses and adopts AI, and you must, must adopt and use AI
in your work to help your constituents and your nation.
Build in requirements around the vendors you use, requirements for what you want to see,
requirements around trust, safety, and security. Encourage and educate your populations about the
use of these tools, and what they ought to demand and seek as consumers. Make sure they understand
the difference between using DeepSeek R1, and your data going to China, OpenAI, and your data going
to, or chat GPT, your own data going to OpenAI, and LLAMA, with some of your data perhaps going
meta, right, but you can also run it open source locally. Who gets the data, what can they do with
it? Make them educated consumers, until they can and should demand systems that bake in trust,
safety, and security. And last, let me just tell you as an investor, we have every incentive to
invest in companies that bake in trust, safety, and security. Why? Because the AI capabilities
that people trust will get deeper adoption. The AI capabilities that are safe and secure
will not only protect their users, and again get more adoption and make them more sticky,
but protect our investment in that intellectual property. Indeed, at Paladin Capital Group,
where we invest in these technologies, we've led a group of two dozen investors who are committed
to investing and putting our investment money to work only in capabilities that are trusted,
safe, and secure. We're also committed to only investing in capabilities that benefit free and
open societies. The United States, our allies, and our partners around the globe, including the
Philippines. And so these things are important, not just to parliamentarians, but to innovators
investors as well. There are opportunities here. Learn from those that have done it right,
and watch out for those ones that those have failed. Thank you. Outstanding. Thank you.
