AI security Is key to U.S. dominance in the AI arms race

Date:
Fri, 28 Feb 2025 15:32:10 +0000

Description:
DeepSeek and security breaches threaten America's AI dominance. Here's how we take it back.

FULL STORY ======================================================================

Donald Trump has the distinction of being the first U.S. President elected in the AI era. While he has rightly recognized the importance of AI with his decisions to appoint a White House AI & Crypto Czar and jumpstart infrastructure investment through the Stargate venture, his administration faces novel threats to U.S. AI dominance. As Trump settles into office, were seeing accelerating competition from new AI models like DeepSeek and
increasing security threats as businesses and critical infrastructure systems integrate AI.

Despite security dangers, it is no exaggeration to say that AI has the potential to be humanitys greatest creation, offering the means to go beyond the bounds of human capability and unlock discoveries that have been out of reach. Whether the U.S. will lead the world in these advancements will be determined within the next four years.

As the leader of the worlds largest economy, Trump will have the ability to influence and guide all of AIs potential into a practical reality. He must focus on two forms of AI security to realize this, addressing both the
security threats AI poses and the need to secure Americas dominance in the AI domain.

Securing AI-driven systems

Sustainable AI innovation cannot occur without balanced protections for its users. Removing barriers to entry to AI is achievable once the burden of safe usage is on the companies creating and implementing AI, not on its everyday users. Organizations should be allowed to freely distribute AI applications
and systems, but only when they also make the associated testing and security reports freely available.

Consider the parallels with access to automobiles: they contain very advanced technology but everyone can see their safety rating and crash-test results before they get behind the wheel. Accessible safety ratings will similarly speed the safe democratization of AI.

In practical terms, smartphone operating systems are the real edge devices
for AI, where most people will access the technology on an hour-to-hour, minute-to-minute basis. Putting the correct legal safeguards in place for how the owners of these operating systems implement AI will have the most
dramatic and immediate impact on AI safety and security.

Meanwhile, the rise of AI agents represents a new vulnerability for U.S. businesses and national infrastructure, as agents can gain access to
sensitive information and run mission-critical tasks such as building and resource control systems. It is all but certain that, in tandem with their
push to innovate their own AI models, countries like China are making significant efforts to hack those of their adversaries. Indeed, there is already evidence of China-based threat actors attempting to exfiltrate information from OpenAI.

With the advent of advanced autonomous agents, the cutting edge will be
agentic warfare, in which these agents will be tasked with finding ways to
hack corporate or national defenses. To win on AI, our security must be as
good or better than our enemies hacking innovations.

Building AI infrastructure

Securing physical infrastructure and the resources required for innovation is also essential to copper-fasten Americas position as the global leader in AI. We can have the smartest companies in the world but we cant win the AI arms race if we dont have the computing and physical infrastructure to stay ahead. DeepSeek has upped the stakes for model cost and performance but the bigger picture is about infrastructure and implementation.

That means reinvigorating our chip manufacturers, investing in
high-performance computing, and making long-term investments in energy infrastructure. Smart trade deals with new and existing partners will ensure Americas access to the GPUs and technologies that underpin AI advances.

Running AI systems requires huge amounts of power: the IEA estimates the electricity use of data centers, AI, and crypto could more than double to
over 1,000 terawatt-hours in 2026 roughly equivalent to the electricity consumption of Japan. U.S. energy supply and grid infrastructure must be up
to the task. Expanding wind, solar, and tidal energy can both power the AI revolution and create new jobs a win-win for America.

Securing U.S. AI dominance

We must also plan for long-term AI innovation. America once led the world in technological research, spearheaded by our 17 National Laboratories that
sprung from federal investment during World War II. The time is right to revitalize those labs, giving them access to abundant AI technology and
funding to staff up and explore this new frontier. The new partnership
between the National Labs and OpenAI is a step in the right direction.

The Trump administration must empower our labs, designating them as places where things can be created, not just homes for long-form research. Instead
of spurring job losses, AI can make jobs safer, more skilled, and
higher-value all good things for our economy. Finding new opportunities for job-specific efficiency is one area where our National Labs can act as hothouses for AI innovation.

Of course, AI can be a powerful tool for government efficiency, with its ability to pull vast amounts of data from various sources and identify
patterns and opportunities for improvements. All this data must be handled delicately, with an extreme emphasis on safety and the security of personally identifiable information. Consider the Department of Veterans Affairs, which has sensitive personal and health information on our 18 million veterans and their families. A data breach there would upset and alienate people and set back AI progress.

Establishing an AI task force

It is clear that the scale of the AI opportunity requires an unprecedented approach that balances appropriate safeguards with speed of action. This can
be achieved by establishing a Department of AI Acceleration, which will allow the U.S. to rip through red tape, apply funding appropriately, and propel American-built AI. Sticking with President Trumps knack for naming things, perhaps AI Force is appropriate?

No matter its title, a dedicated agency that allocates funding to those developing AI in responsible ways will achieve better results than time-consuming regulation and legislation. If we want the US to accelerate
and dominate in the AI space, lets assemble the greatest minds in a single place, with the sole purpose of ensuring safe AI that benefits everyone.

Quickly establishing this landmark new department, backed by targeted investment in research and infrastructure and an agreed approach to AI security, will make America the global authority on safe AI not just leading the way but lighting a path forward for the rest of the world. The
opportunity is ours for the taking.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry
today. The views expressed here are those of the author and are not
necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

======================================================================
Link to news story: https://www.techradar.com/pro/ai-security-is-key-to-u-s-dominance-in-the-ai-ar ms-race

$$
--- SBBSecho 3.20-Linux
* Origin: capitolcityonline.net * Telnet/SSH:2022/HTTP (1:2320/105)