LUKE FELTHAM | How tech billionaires became dogs of war

Sam Altman has had a busy week. The OpenAI chief executive opportunistically swooped in on a deal with the Pentagon, was forced to acknowledge he did exactly that, faced further backlash, and eventually promised that the appropriate guardrails had been added to the deal — mollifying precisely no-one.

The saga began when a $200m deal between the US Department of War and Anthropic fell apart. The latter had demanded assurances that the military would not use its Claude model for either domestic surveillance or autonomous lethal weapons systems. Anthropic walked away when they did not materialise. US President Donald Trump said he “fired them like dogs” and designated them as a supply chain risk — effectively a blacklisting.

But users rewarded the red line: Claude has spent the last week on top of the Apple App Store.

ChatGPT, meanwhile, whose owners OpenAI happily picked up the pieces of the deal, has faced reports of rapid uninstalls. (Although, in fairness, it currently sits second in the same store).

Altman was reflective, bordering on repentant: “We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy.”

His promise is that his software will not be used for domestic spying. As for the part about being used with or in conjunction with deadly weaponry… well, we can infer that the gun manufacturer doesn’t control who its product is pointed at.

The last year has been so numbing that it’s easy to forget how unthinkable this sequence of events would have been in the near past.

Gone are the days of the geeky caricature of the billionaire tech CEO. He was a simple, innocent creature. With a clumsy smile stapled to his cheeks, he stumbled onto stage with the charisma of a hostage with a knife at their back.

Those days are over. Today’s tech leaders are intertwined with the military-industrial complex on a scale that was not just previously unthinkable but effectively forbidden in Silicon Valley.

For all its flaws, the land of venture capitalists and unicorns has always been wedded to a pacifist philosophy. If 21st-century tech inherently promised a progressive, utopian world, violent conflict was a non-starter.

But that all changed when it became clear that Trump would be re-elected (which looked probable months before November 2024). His inauguration the following January was a rare moment when we were all palpably aware that history was being written in real time.

The company moguls that filled the front rows represented every major platform our global society uses to search, consume and share information. There was no longer value in being an iconoclast boasting of being diametrically opposed to the halls of power.

Whereas Altman is squeamish about how he publicly presents the department of war deal, Karp displays no such compunction in his own dealings, happily acknowledging that his tech is ‘on occasion used to kill people’.

Meta founder Mark Zuckerberg quite literally embodied the shift in attitude. He dropped his staid, stick-in-the-rear persona and reinvented himself as a combat sport-obsessed, freedom-of-speech acolyte. He proclaimed that his product had gone astray from the network of human connections and communication he had first envisioned. (Few will forget that the first iteration of what was to become Facebook was actually created to rate college women.)

Amazon’s Jeff Bezos’s compliance manifested in equally obvious ways. The Washington Post owner went from sternly defending the publication at great personal financial risk to ordering it to break from the 36-year tradition of backing a presidential candidate. More than 250,000 people dumped their subscriptions, undoubtedly one of the factors that saw 300 jobs cut last month.

But there is one man above all others who represents — and is actively pushing for — Silicon Valley’s new ideology.

Alex Karp has turned Palantir into one of the most valuable companies in the world. Its intelligence services boast a vast client list that includes the US military, the Ukrainian military and the Israel Defense Forces.

Whereas Altman is squeamish about how he publicly presents the department of war deal, Karp displays no such compunction in his own dealings, happily acknowledging that his tech is “on occasion used to kill people”.

His words for the Anthropic-OpenAI fiasco were similarly abrasive.

“If Silicon Valley believes we are going to take away everyone’s white-collar job… and you’re gonna screw the military — if you don’t think that’s gonna lead to nationalisation of our technology — you’re retarded,” he said on a panel last week.

Modern warfare is utilising AI and other tech in terrifying ways.

A remarkable piece in the Financial Times last week (“Inside the plan to kill Ali Khamenei”) detailed how Israeli intelligence hacked into street cameras in Tehran and modelled complex algorithms to nail down the location of Iran’s head of state. The information enabled the US to start the war on its terms.

Iran used similar forms of espionage to gain eyes in Tel Aviv during the 12-day war last year.

The geopolitical technology race is as alive as it has ever been. And it is creating profits that can’t be turned away from.