The Column – Cyber Security Week
In 1963, at the height of the Cold War, the British government was overwhelmed by a story that had not needed bugging. John Profumo, Minister of War, had had an affair with Christine Keeler, who was also dating Yevgeny Ivanov, Soviet naval attaché in London. The fear that confidential information would reach Moscow aggravated the scandal, along with Profumo’s lies in front of the House of Commons. The minister resigned and Lord Denning’s inquiry did not find any compromise to national security, but the political damage had already been done.
Sixty-three years later the setting has changed, but the script retains something familiar. At the center of the story is Jensen Huang, founder of Nvidia, an artificial intelligence giant, who resigned after revelations about some personal relationships and suspicions regarding Alice Huang, former manager of a Singapore technology company, suspected of links with Chinese intelligence. The coincidence of the surname does not indicate any kinship, just as the suspicions do not constitute evidence of espionage: no public evidence demonstrates that Huang has transferred secrets to Beijing and Nvidia has declared that it has no elements to believe its confidential information has been compromised.
The comparison with Profumo does not imply equivalent responsibilities, but it shows how easily a personal relationship can cross the border between private life and strategic power. In 1963 it was feared that a minister’s confidences would reach the Kremlin; today a relationship takes on a different weight when it concerns those who produce the processors essential to artificial intelligence. There is no need to imagine a dossier secretly stolen, because knowing commercial orientations and future decisions can be of great value, especially when those choices affect the balance between the United States and China.
Technology has certainly changed the scale of the problem. The manager of a digital giant participates in negotiations, directs investments, meets governments and decides which markets to control; the more power is concentrated in a few organizations and in the hands of a few individuals, the greater becomes the value of their information, but also of their weaknesses, real or perceived. Here’s the paradox: we protect machines with increasingly sophisticated encryption and controls, while desire, ambition, vanity and the need for trust continue to escape any software update.
However, it would be wrong to transform the private lives of managers into a matter of permanent surveillance, as if a romantic relationship in itself constituted a threat. The problem arises when a personal bond meets a position on which decisions of collective importance depend and there is a lack of tools to recognize and govern any conflicts of interest. We can enforce two-factor authentication to access a corporate network, but applying it to feelings would be quite complicated; it is more reasonable to distribute responsibilities and set up counterweights capable of functioning even when people prove fallible.
The so-called human factor is often evoked when thinking about the employee who opens a suspicious attachment or chooses an embarrassing password, almost as if it were a problem that can be solved with a training course and a few warnings. However, there is another dimension, less convenient to deal with, which concerns those who concentrate power, information and relationships: here vulnerability is not necessarily a technical error, but rather the difficulty of completely separating the person from the role they cover. This applies to technology companies as well as governments, strategic infrastructures and security systems.
The Profumo scandal did not prove that a secret had reached Moscow and the Huang case, at least as far as is now publicly ascertained, does not demonstrate that a secret had reached Beijing. Both remind us of how thin the boundary is between what we consider private and what, due to the possible consequences, becomes public. We have replaced teleprinters with machines capable of processing unthinkable quantities of data, but an equally revolutionary update does not seem to have been distributed regarding the functioning of people. Chips change with every generation. We much less.




