Artificial intelligence is rapidly becoming more than a technology story. It is becoming a geopolitical story — and increasingly, a story about who gets to define what the world should fear.
China has now pushed back against calls for a slowdown in advanced AI development, accusing those promoting confrontation and alarm of engaging in “fearmongering” and “vicious competition.”
The timing is significant.
The remarks came after Anthropic CEO Dario Amodei argued that the United States and other countries should consider slowing the development of frontier AI, while also warning about the consequences of China gaining an advantage in advanced artificial intelligence. OpenAI CEO Sam Altman and Elon Musk have also expressed support for greater caution around frontier AI development.
China's response was essentially: do not turn legitimate AI governance into another technological Cold War.
Chinese Foreign Ministry spokesman Guo Jiakun said China supports an open, inclusive and beneficial approach to AI, warning that fearmongering, confrontation and vicious competition could damage global AI governance.
And this is where the AI debate becomes much more complicated.
Fear can be a political technology
There are legitimate reasons to worry about advanced AI.
Systems are becoming increasingly capable of writing software, conducting research, operating as agents and interacting with digital infrastructure. Governments and companies are already examining risks involving cyberattacks, misinformation, autonomous systems and the possibility that increasingly capable AI could become difficult to control.
Reuters reported on Monday that China itself is preparing for scenarios involving AI systems potentially escaping human control. Chinese officials have also been developing regulatory frameworks aimed at autonomous AI systems and emphasizing the need to keep powerful AI under human control.
So Beijing's position is not simply that AI has no risks.
That would be difficult to defend.
The more interesting argument is that the language used to describe those risks can itself become a geopolitical weapon.
If one country says, “This technology could become dangerous if our rival develops it first,” the conversation can quickly move from safety to strategic competition.
And once that happens, every AI safety warning can be interpreted through the lens of national power.
The China problem is changing the meaning of AI safety
Consider the argument surrounding Chinese AI.
Amodei's position is not merely that AI could become dangerous. He has warned that allowing China to pull ahead in advanced AI could have serious geopolitical consequences and has supported restrictions intended to limit China's access to advanced AI chips and technology.
That creates an obvious tension.
If the United States tells China:
“AI is so dangerous that humanity must slow down.”
But simultaneously says:
“We must prevent China from obtaining the technology because we cannot allow China to win the AI race.”
China has every incentive to question whether the argument is genuinely about AI safety.
Chinese state media has already portrayed the slowdown argument as an attempt to widen America's technological advantage before negotiating from a position of strength.
Whether that interpretation is fair is another question.
But it is not difficult to understand why Beijing would be suspicious.
The irony: AI fear is coming from inside the industry
There is another uncomfortable part of this debate.
Some of the loudest warnings about AI are not coming from people who hate artificial intelligence.
They are coming from the people building it.
Amodei runs Anthropic, one of the world's major frontier AI companies. Altman runs OpenAI. Musk has repeatedly warned about AI risks while simultaneously investing heavily in AI.
That makes it difficult to simply dismiss every warning as anti-technology hysteria.
But it also raises another question:
How much of AI safety discourse is genuinely about protecting humanity, and how much is about controlling the competitive environment in which AI companies operate?
That question deserves to be asked without automatically assuming either side is right.
The AI industry has an incentive to create fear — and an incentive to create optimism
This is where the public needs to become more sophisticated.
AI companies have enormous financial incentives to promote the transformative potential of their technology.
The bigger the promised economic revolution, the easier it becomes to justify enormous investments in computing infrastructure, chips, data centres and AI research.
But there is another incentive.
If the technology becomes politically associated with catastrophic risks, the companies developing it can potentially gain greater influence over regulation.
The people who understand the technology best can position themselves as the people best qualified to decide how it should be controlled.
That does not mean AI companies are deliberately inventing risks.
It means society should distinguish between a real technological risk and the political narrative constructed around that risk.
China is not innocent of AI fear either
There is an important irony in Beijing's accusation of fearmongering.
Chinese officials have themselves warned about AI being used for cognitive warfare, political manipulation, deepfakes, cybersecurity attacks and threats to national security.
China's State Security Minister Chen Yixin has reportedly highlighted multiple AI-related risks, including threats to political stability, infrastructure and national security.
So the argument is not really:
China says AI is safe while America says AI is dangerous.
It is more complicated.
Both countries recognize that AI can become dangerous.
The disagreement is increasingly about who controls the technology, who establishes the rules and whose definition of “security” becomes the global standard.
The danger of turning AI into another arms race
This may be the most important part of the argument.
If Washington believes China must be slowed down, while Beijing believes Washington is trying to preserve technological dominance, both sides have an incentive to accelerate.
And that creates precisely the dynamic AI safety advocates claim they are trying to prevent.
Imagine two countries developing increasingly powerful AI systems while each fears that slowing down unilaterally will allow the other country to gain an advantage.
That is essentially a technological version of an arms race.
The logic becomes:
We cannot slow down because they might not slow down.
And then:
They cannot slow down because we are not slowing down.
The result is acceleration without trust.
That is why the forthcoming US-China discussions about AI matter. The two countries have been preparing for AI-related dialogue as concerns over increasingly capable frontier systems grow.
But calling everything “fearmongering” is dangerous too
There is a trap on the other side.
If every warning about AI is dismissed as fearmongering, governments and companies could ignore genuine problems until they become much harder to manage.
AI-generated misinformation is real.
Cybersecurity risks are real.
Deepfakes are becoming increasingly difficult to detect.
Autonomous AI systems create new questions about accountability.
And the possibility of AI systems behaving in ways their developers did not anticipate cannot simply be laughed away.
Recent research has even highlighted how difficult it can be for humans and existing detection systems to distinguish sophisticated AI-generated video from authentic footage.
Therefore, the answer cannot be:
“Stop talking about AI risks.”
Nor should it be:
“Stop developing AI.”
The better question is:
Which risks are measurable, which are speculative, and which are being amplified for political or commercial purposes?
Africa should be paying attention
For countries such as Nigeria, this debate should not be treated as a distant argument between Washington and Beijing.
AI governance will eventually affect everyone.
If the world's two largest AI powers establish the rules, developing countries may find themselves adopting standards they had little role in designing.
That matters for data.
It matters for national security.
It matters for education.
It matters for digital infrastructure.
And it matters for sovereignty.
African countries should therefore resist becoming passive consumers of whichever AI ecosystem wins the geopolitical contest.
The continent needs its own conversations about data ownership, AI infrastructure, local-language models, cybersecurity, public-sector AI and the rules governing foreign AI companies operating within African markets.
Otherwise, the debate over AI safety could eventually become another debate in which powerful countries decide what is safe, what is dangerous and what developing countries are allowed to access.
The real fear should not be AI alone
There is a tendency to frame the future as a choice between two extremes.
Either AI will save humanity.
Or AI will destroy humanity.
Reality is likely to be considerably less cinematic.
AI will probably create enormous opportunities while simultaneously creating new forms of inequality, surveillance, misinformation, cybercrime and economic disruption.
The biggest danger may therefore not be an evil machine.
It may be humans using fear about machines to justify political decisions that would otherwise be difficult to sell.
That is why China's accusation of “fearmongering” deserves examination — even if Beijing's own geopolitical motives deserve scrutiny.
The world should not dismiss AI safety concerns simply because China raises the issue.
But neither should the world accept every catastrophic prediction simply because it comes from a prominent AI executive.
AI needs regulation. AI needs safety research. AI needs international cooperation.
But it also needs something that is becoming increasingly rare in the technology debate:
intellectual honesty.
If a risk is real, demonstrate it.
If a risk is speculative, label it as speculative.
If an argument is about national security, say so.
And if a safety argument also happens to strengthen one country's position against another, the public deserves to know that too.
Because the greatest mistake the world could make in the AI era is allowing fear itself to become a tool of technological competition.
Comments
Post a Comment