Autonomous mobility and the lessons of the competition in artificial intelligence
In February 2025 I read this article about the potential risks of recent developments in and it gave me a lot to think about.
The race to reach AGI (artificial general intelligence) and gain an edge over the other competitors may be making us set aside considerations we should never forget, given what is at stake and the possible threats that misuse of this kind of model can bring.
In this regard, Sam Rubin, senior vice president at Palo Alto Networks, said that DeepSeek's new model is more vulnerable to possible 'jailbreak' techniques.
Is this just criticism because the company that developed it is Chinese? I don't think so, but it is also important to remember that these risks exist in other models, although perhaps they are better controlled.
I myself have easily managed to give ChatGPT and Claude instructions the models had not been programmed for. This, although useful and, I must admit, very desirable on many occasions given these companies' arbitrary restrictions, is at the same time hugely worrying because of the threats that can arise from it.
Cybersecurity lessons from connected autonomous mobility
As you know, which is why you are here, at AEVAC we work in the world of connected autonomous mobility. And if there is one thing we can say, it is that this sector has been building cybersecurity considerations into every aspect of autonomy and connectivity for years to minimise risks from the start.
However, seeing the mistakes being made in releasing LLMs and the ever greater, ever more aggressive competition in AI, I wonder whether, despite all the ground covered, something similar could happen in the autonomous driving sector.
The result could be a race in which technological vulnerabilities put at risk the work done over all these years, creating public distrust and holding back the investment that is needed.
Smart regulation: balancing innovation and safety
Those who know me know I am a convinced tech optimist. I believe in continuous innovation as a wonderful tool for improving our quality of life. However, radical changes need a more responsible approach than the one we are seeing with some models.
I am not in favour of excessive regulation, but in this case I do think smart involvement by the authorities is necessary, which, to be honest, I am not seeing, at least in Europe. We need more , experimentation and collaboration between the public, private and third sectors to move faster, go further and do so safely.
I am not talking about censorship, in LLMs or in the future of autonomous mobility. I am only talking about intelligence, about awareness of what we are playing for, for better or worse.
As a user of public roads, I demand that when a vehicle goes out on the street, it does so with the highest safety guarantees, so that we can all trust them, so that we all know how to behave around them and enjoy their full potential.
And as citizens, it is our duty to demand progress and also safety. We cannot look the other way or allow our future to be cut short, by putting obstacles in the way of progress or ignoring the possible risks. As institutions and companies, it is their duty to make it happen.
Share this article
- Share on LinkedIn (opens in a new window)
- Share on Facebook (opens in a new window)
- Share on X (opens in a new window)
The social network links load nothing from them until you click. They open in another window.