AI and geopolitics: A modern Cold War unfolds

28 January 2025

AI-Enhanced image processing and computer vision
Dr Mahdi Aiash - profile photo

Article Written By

Dr Mahdi Aiash

Middlesex University expert speaks about the potential security risks and global implications amid the rapid rise of China’s DeepSeek AI platform

The rapid rise of DeepSeek, a Chinese AI company founded in 2023, is making waves in the tech world and beyond, according to Dr Mahdi Aiash, an Associate Professor of Computer Science and Cyber Security at Middlesex University. He also leads the Cybersecurity Research Group at the University.

In this article, Dr Aiash explains how by early 2025, DeepSeek had already positioned itself as a major player in AI, garnering global attention for its cutting-edge R1 model. But Dr Aiash says that DeepSeek’s story isn’t just about technology - it’s about power, influence, and a growing sense that the global AI race is turning into a modern-day ‘Cold War’.

A new ‘cold war’ in AI advancement

When DeepSeek announced its R1 model, it turned heads across the globe. Built using pure reinforcement learning, this system rivals some of the most advanced AI models out there, including those from OpenAI. What really shocked the industry, though, was their decision to open-source the model under an MIT license. It’s a bold move that directly challenges the proprietary approach of many Western tech companies.

DeepSeek’s rise to prominence, exemplified by its 2025 launch of the R1 model, highlights a broader shift in the AI landscape. This development isn’t just a testament to China’s growing technological expertise - it’s a clear signal that AI has become a strategic tool in global geopolitics. As nations like China push forward, the balance of power in technology-driven influence is evolving rapidly.

Adding to these concerns, recent reports, such as this one from The Guardian, suggest that advanced AI technologies, like those developed by DeepSeek, may already be influencing the cybersecurity landscape in significant ways. While it remains unclear who is responsible, the potential use of AI in orchestrating complex cyberattacks highlights the growing stakes in the global AI race.

The regulatory divide: europe and the UK vs. US and China

Here in the UK and across Europe, we’ve been focused on regulating AI - and for good reason. As we highlighted in our earlier blog, “What are the AI Safety Regulations amid Launch of ChatGPT-4o?”, these rules are crucial for addressing safety and ethics. However, there’s a catch: regulations can significantly slow down innovation. While we’re busy debating the finer points of ethics, the US and China are surging ahead, taking more risks with fewer restrictions.

A prime example of this came just days ago, when President Donald Trump announced the “Stargate” initiative, a collaboration with private-sector giants like OpenAI, SoftBank, and Oracle. This ambitious program aims to inject up to $500 billion into AI infrastructure over the next four years, with an initial investment of $100 billion to establish advanced data centres across the US. This massive commitment highlights America’s aggressive approach to staying ahead in the AI race and contrasts sharply with Europe’s slower, regulation-first strategy.

In China, the government’s heavy involvement in AI acts like a booster rocket for development. In the US, private companies are pushing the boundaries, even as regulators struggle to keep up. The UK’s more cautious, “regulate first, innovate later” approach risks leaving us behind. The challenge is finding a way to stay competitive without compromising on safety.

Game-changer: AI in defense and the supply chain question

AI isn’t just about chatbots and smarter apps - it’s shaping defence strategies too. Take the Lavender model, for example, which Israel deployed during its conflict with Lebanon. As this article highlights, Lavender played a pivotal role in shaping the war’s outcome by providing real-time insights and speeding up decisions. It’s proof of how much AI can change the game in high-stakes scenarios.

This brings up a big question: who controls the AI supply chain? Relying on foreign tech could leave nations vulnerable, especially in moments of crisis. For Europe and the UK, ensuring supply chain security is a challenge that can’t be ignored. As AI becomes more integrated into critical operations, the risks of dependency on external providers - especially those linked to geopolitical competitors - grow exponentially.

AI in the automotive industry and Gulf expansion

AI is also disrupting industries like car manufacturing. For example, Chinese automakers exported more than 3.2 million vehicles in 2024 alone, outpacing many Western competitors. By using AI to design smarter, more efficient vehicles, Chinese automakers are gaining significant market share, often at the expense of German, US, and other Western brands.

Meanwhile, China is making serious inroads into Gulf markets. By offering high-tech yet affordable alternatives to Western products, Chinese companies are cementing their position in these lucrative regions. This isn’t just an economic trend - it’s a shift that could see China becoming the global leader in innovation, further threatening the West’s dominance in education and research.

Security concerns: the risks of adoption

There’s another side to all this progress: security. Companies like DeepSeek operate in a country known for using technology for state surveillance. Everyday gadgets, like internet-connected doorbells, have reportedly been repurposed for monitoring. In my view it’s no wonder people are hesitant to adopt Chinese AI platforms.

This isn’t the first time Chinese technology has sparked global concerns. Take TikTok, for example  - a wildly popular social media platform that’s come under scrutiny in the U.S. over fears it could be used to collect data for the Chinese government. Discussions around potentially banning TikTok have highlighted the unease many countries feel about using Chinese-developed technologies, particularly those handling sensitive data.

Beyond surveillance, there’s the worry of sensitive data falling into the wrong hands. For many organizations, these risks are a major hurdle - no matter how advanced the technology might be.

AI for cyberattacks: a silent threat

AI is increasingly being leveraged as a tool for cyberattacks, presenting one of the most concerning aspects of its development. Imagine an advanced AI model that potentially will become capable of automating and coordinating large-scale cyberattacks across critical systems worldwide. These tools could adapt to defences in real-time, escalating the scale and precision of the damage they inflict. Worse still, these tools could be used by China’s allies to launch attacks while concealing Beijing’s direct involvement. This kind of proxy strategy makes accountability nearly impossible and complicates international responses.

Automated advanced persistent threats (APTs) present another alarming scenario. Traditionally requiring skilled human attackers, these operations could now use AI systems to conduct prolonged, stealthy intrusions, collecting data or disrupting systems with minimal oversight. Similarly, AI-driven systems could mask the origin of cyberattacks, enabling deniable operations against rival nations and eroding trust in digital ecosystems.

AI also supercharges cyber espionage. By automating the collection, analysis, and weaponization of stolen data, adversaries can operate at unprecedented speed and scale. The situation becomes even more concerning with the deployment of platforms like social media apps, such as TikTok, which can provide vast amounts of user data essential for AI performance. These platforms inadvertently act as data goldmines, feeding AI models with the behavioural, geographical, and social patterns needed to enhance cyber operations. This could overwhelm traditional cybersecurity defences, leaving governments, businesses, and individuals vulnerable to highly efficient and relentless attacks.

Beyond TikTok, other technologies like Huawei’s 5G infrastructure and advanced facial recognition systems have raised similar concerns about data security and potential misuse for state surveillance.

Moving forward: what’s next for the UK and Europe?

The rise of DeepSeek sends a clear signal: the global AI race is accelerating, and for the UK and Europe, it’s time to reassess our approach. Striking a balance between fostering innovation and maintaining robust regulation is no small task, but it’s essential if we want to stay competitive and avoid falling behind.

At the same time, we can’t afford to ignore the mounting risks. From surveillance concerns and cybersecurity threats to losing ground in education and research, the challenges are multi-faceted and growing. Addressing these issues requires more than internal efforts; it demands global collaboration and the establishment of international norms.

A critical gap remains in the realm of AI governance, particularly in cyber warfare. Despite the growing threat, there are no robust international agreements regulating the use of AI for offensive operations. This regulatory vacuum poses a significant challenge, making it harder to hold bad actors accountable and mitigate risks at a global scale.

To secure a leading position in this rapidly evolving landscape, the UK and Europe must prioritize investment in AI innovation, technology security and safety, push for international cooperation on governance, and foster strategic partnerships between governments and the private sector. The stakes are high - are we ready to adapt and lead, or will we allow others to shape the future of AI?

References

Continue your journey