Reading · In the future, everyone will be famous for 15Mb

Dateline: Manchester, 2nd July 2026.

The almost inevitable news that Meta is building a “photorealistic” AI-powered 3D Mark Zuckerberg will surely set in motion an unstoppable trend for Silicon Valley CEOs. Your own personal Elon Musk is just around the corner. But how will know you whether you are being fired by the real fake CEO or a fake fake CEO?

Mr. Zuckerberg is apparently personally involved in training and testing his avatar to engage in conversation with employees, programming the models behind the avatar with his mannerisms and thinking on company strategies. The idea is that employees might feel more connected to their boss by meeting him in the metaverse.

Whether they will nor not, I couldn’t say, but it is interesting to see him follow the trail blazed by noted pop personality FKA Twigs. She testified to a US Senate Judiciary subcommittee that she had developed her own deepfake version of herself (trained in her personality and able to speak in French, Korean and Japanese) so that she can leave the bots to interact with journalists and her many fans while she focuses on her music. It was great to see this farsighted and innovative use of new technology coming from a British creative and I was not surprised to see Mr. Zuckerberg following down the same path.

(I’ve often said that artists can tell us more about the future than technologists. This is why, for example, I am the sponsor of the Future of Money Design Awards for art students that were presented at Pay360 in London earlier this year.)

Ms. Twigs said that in an age that demands a lot of PR and promotional work, her AI double will free her up for art and I am sure that Mr. Zuckerberg’s AI double will similarly free him up for more creative tasks than talking to employees. They are not the only ones treading this path. Reid Hoffman (a co-founder of LinkedIn) follows the Twigs trail and uses his digital twin, Reid AI, for public appearances and media interviews. His system (trained on 22 years of his books, speeches, podcasts and articles) has delivered more than 75 addresses and presentations since its 2024 launch.

(Both Klarna’s Sebastian Siemiatkowski and Zoom’s Eric Yuan have also used AI-generated versions of themselves to deliver portions of earnings communications.)

I should say that much like FKA Twigs, I have a deepfake version of myself up and running so that I can have more time to focus on Dungeons & Dragons. My digital twin, DaveGPT, has been gainfully employed since 2023. I’m thinking of adding DaveGPT access to my OpenClaw bot to make him available to the masses 24/7/365, but I remain concerned about the legal implications of hallucination in a public place.

(Although, I cannot help but note, lawyers seem to not share such concerns. Yet another top law firm has just been reprimanded by a judge for using AIs to generate slop presented in court. Pinsent Masons misled a court twice by inaccurately citing a statute in relation to a routine insolvency application, an error that was only picked up when the judge queried the reference.)

The trend toward AI CEOs is surely unstoppable. Apart from anything else, as a study in Harvard Business Review that benchmarked AIs performance against four human participants (two students and two executives) found, AI consistently outperforms top human participants on nearly every metric and in a business simulation surpassed the best-performing student’s market share and profitability three rounds ahead.

Incidentally, given my morbid nature, I wonder if listening to employees complaining about the flavours of free ice cream in the cafeteria is really the main reason for the Silicon Valley Dupes. I think there may be a much more serious reason for CEOs to retire to their bunkers and send their digital doppelgängers (digigängers?) out into the world. We have already seen one CEO assassinated on the way to a board meeting in New York and a Molotov cocktail thrown at the house of OpenAI CEO Sam Altman.

I wonder if there is a nascent Neo-Luddite revolution in the offing?

I say this because, contrary to the use of “Luddite” as a synonym for technophobes, those original loom wreckers were not opposed to technology in the abstract. They were concerned about some specific uses of machinery that they thought would de-skill their trade, undercut wages and shift bargaining power dramatically towards mill owners.

As it turned out, they were right.

The Industrial Revolution meant rising aggregate productivity and national income, but with severe, uneven short‑term impacts on particular groups. The Luddites correctly saw that control over technology would be used to restructure power relations: ownership of capital‑intensive machinery allowed employers to replace independent or semi‑independent craftsmen with regimented wage labour under worse conditions. In that sense, their campaign was as much about resisting a new labour regime and loss of autonomy as about preserving an old technology.

(Remember that the later social advances for industrial workers—unionisation, labour law, social insurance—came not from stopping technology but from political organisation around its distributional consequences and the same will be true for AI.)

Now, it is fair to observe that the AmericanNeo-Luddites reacting against Silicon Valley’s AI, data centres and cryptogrifting have access to technologies that their English ancestors did: social media and automatic weapons, for example, will make for a very different kind of uprising that might prove more resilient that artisans with hammers did.

So with a seeming backlash against Big Tech in progress, the spectre of Neo-Luddite paramilitaries armed with automatic weapons abroad and with prediction markets booming in the darkest corners of the internet, it is hardly surprising that spending on CEO security is shooting up.

In the US, median spending on security for C-level executives had already gone up 16% to a record $106,530 in 2024 with coverage going from a quarter of those executives to a third. Technology companies had the biggest growth in implementing security measures for executives, with a three-quarters jump in those receiving the benefits, and that was long before someone opened fire on Sam Altman’s house (two days after the Molotov cocktail attack).

Sending your AI to testify before Congress in the metaverse or tasking your avatar with visiting employees in their cubicles to brighten their days, while you remain in an underground lair below a Pacific island, would seem the natural modus operandi for a happening 21st-century corporation.

This is all well and good, and in time we may all find ourselves preferring to deal with digital CEOs rather than the real ones, but in an era of rampant fakery, I do wonder how employees, investor, business partners or indeed politicians will know that they are looking at the real fake Mark and not a fake fake Mark?

The problem of bogus AI-powered CEOs is much in my mind as I write, by the way, because a good friend of mine was recently subject to a clever fraud attempt that used a voice clone of his real CEO via WhatsApp. The fraud was averted at the last minute—hat tip to the JPMC fraud team—but it came pretty close to fooling a very smart person. These attacks are global, relentless and getting better all the time. You have been warned.

Give a gift subscription

We know what is needed (e.g., digital signatures) but not how to make it part of the infrastructure. We will get nowhere trying to explain public key cryptography to the general public: we need to make it happen for them. I think the idea of some basic red/yellow/green signalling should be explored. If the call comes from the CEO, then the CEO’s name should show up in green on your phone. If the call comes from a real person, but the system cannot tell if it is your boss, then it should show up in yellow. If the call doesn’t come from a real person, then it should show up in red. We need both the effective cryptographic tools to defend against the deepfakes and the UX/CX to actually help people to survive the attacks.

So, if the real fake Mark Zuckerberg shows up in your cubicle to fire you then the flashing red/green signals on your phone would alert you to the fact that you are not dealing with a real person but you are dealing with an agent acting with Mark’s authority. Quite simple.

Faking Real Fake Leadership

Comments