Showing posts with label s4. Show all posts
Showing posts with label s4. Show all posts

Monday, August 10, 2026

17563: On S4 Capital H1 P&L OMG WTF BS.

 

MediaPost spotlighted the S4 Capital H1 2026 report, indicating the White holding company continued to experience net revenue declines. Dramatic cost cutting, however, helped to ignite a profit boost.

Gee, Sir Martin Sorrell might earn the distinction of being unable to orchestrate a financial turnaround for the biggest White holding company (WPP) and his current peanut factory.

That’s quite a range of abject failure.

S4 Shares Soar 26% On First-Half Profit Boost

By Steve McClellan

Martin Sorrell-led S4 Capital continued to shrink in the first half of the year with declines in reported and organic net revenue while sharp cost cutting led to a profit boost. The firm also declared a dividend and reduced debt during the period. Shareholders applauded, giving S4 shares a 26% bump up in Wednesday trading after the release of the firm’s first-half results.  

Reported net revenue for the first half was 308 million GBP (approximately $415 million), down 6.2% with a 4.7% organic net revenue shortfall.  

S4 sited continuing macroeconomic uncertainty exacerbated by the Middle East conflict as part of the reason for the revenue falloff. Also, some clients spent less with the firm while boosting capital expenditures in artificial intelligence infrastructure. “Clients continue to be cautious leading to longer sales cycles,” S4 stated. 

But pre-tax profits were up 82.7% to a record 38 million GPB ($51 million). Cost reductions included back-office efficiencies and staff cuts. Total staff at the company was down 10.5% as of June 2026 versus a year ago to 6,150.  

The firm downgraded its organic revenue outlook for the full year, forecasting a decline in the mid-single digits versus the previous “slight dip” the company guided to at the end of the first quarter. But pre-tax profits should reach the analyst consensus of 85 million GBP ($115 million) with a profit margin increase of 1.4%.  

“We anticipate that clients will remain cautious in the near term reflecting heightened macroeconomic uncertainty, including the continuing conflict in the Middle East,” stated Sorrell. “While the macroeconomic environment remains uncertain, we see growing opportunities as clients become more selective about growth geographically and increasingly focused on implementing technologies such as AI, Blockchain and Quantum to drive efficiency.”  

The Americas, the firm’s largest region by revenue, was down slightly (0.8%), while Europe and Asia Pacific were both down double-digits.   

The company’s marketing services unit posted net revenues of 281.9 GBP ($380 million), down 4.4% organically while technology services totaled 26.1 million GBP ($35 million), down 7.4%.

Monday, June 15, 2026

17508: Sir Martin Sorrell Surveys His Failures.

Adweek interviewed Sir Martin Sorrell, who vomited vitriol in standard Sorrellian style.

 

The trade journal stated, “Sir Martin Sorrell has been an expert on M&A for almost half a century.”

 

Has he? It could be argued Sorrell’s expertise seriously eroded over the past decade at least.

 

His S4 Capital venture seems closer to breakdown than breakthrough.

 

Meanwhile, Sorrell’s notorious achievement—WPP—is a flaming dumpster. Indeed, the White holding company was death-spiraling toward a catastrophic crash landing long before Sir Peanut pulled the ripcord of his golden parachute.

 

Sir Martin Sorrell has been an expert…? At this point, he’s an expert has-been.


Sir Martin Sorrell Sees No Easy Exit for Holding Companies

 

Accenture should buy WPP and more insights from the S4 Capital executive chairman.

 

By Ryan Joe and Alison Weissbrot

 

Sir Martin Sorrell has been an expert on M&A for almost half a century. He founded WPP through reverse acquisition in 1985, and built it up into a towering ad giant that snapped up major agency brands like J. Walter Thompson, Ogilvy & Mather, and Young & Rubicam. 

 

Following an ignominious departure in 2018—WPP accused him of misusing company assets—Sorrell plotted a comeback, founding S4 Capital just weeks later. 

 

S4 built itself up through acquisition, notably production agency MediaMonks and performance agency MightyHive, positioning itself as a digital-first holdco. Lately, that thesis has been severely tested as many tech clients, where S4 has heavy exposure, pulled back on marketing spend to invest in AI. 

 

Revenue contracted, and shares fell about 38% over the course of last year, though its early 2026 results have indicated improvement. S4 currently views itself as being in a period of stabilizing its business before revisiting growth-by-acquisition. 

 

ADWEEK sat down with Sorrell to get his read on the market, who should buy whom, his competitors (whom he’s never been shy to criticize), and S4’s next steps.

 

This interview has been edited for length and clarity. It was conducted before Publicis said it would buy LiveRamp, and Accenture said it would buy Whalar.

 

ADWEEK: The ad agency holdcos have gone from the biggest industry acquirers to acquisition targets. What’s the outlook?

 

SIR MARTIN SORRELL: There are very few, if any, exits. No activists have stepped in.

 

There are no takers. Most of the big deals would involve syndicates. One PE firm couldn’t write the check for WPP or Dentsu. Maybe Bain Capital could have done it on their own.

 

Besides Publicis, is any big holdco a buyer at this point?

 

Omnicom probably isn’t. Dentsu is not going to be in buying mode. Stagwell will try. The irony may be, this is the time to take the risk because valuations are so battered. But I don’t think people would take a risk.

 

If anyone was going to take a risk, who do you think it would be? 

 

Well, Havas. Maybe not of scale. But [Havas CEO Yannick] Bolloré always sends you in the wrong direction. Always. Keeps you guessing.

 

And I think Accenture. If I was [Accenture CEO] Julia [Sweet], I would go for WPP. With them, they could get significantly better. I think the media piece in their hands would be very valuable. But whether they’ve got the balls though… It would be messy. But you’ve got David Droga there.

 

Publicis has been on a successful run. Is it sustainable?

 

What Publicis does—everybody says it’s all proprietary trading, which is why I think The Trade Desk thing is dangerous territory for them, they’re going to turn the microscope on themselves—is they look at the client as a whole.

 

Omnicom and WPP are falling into the trap of the matrix being driven by capability. The reorganization of WPP and the reorganization of Omnicom is capability-first, client second, geography third. Publicis and, indeed, ourselves is geography first, client second, capability third.

 

What’s wrong with organizing around capability?

 

You have warring factions. If you do it by country, you have inter-country rivalry, inter-region rivalry, that’s true, but you get people to integrate what they do. Publicis got it right. Can they maintain it? Well, the industry wisdom is these things go in cycles. 

 

Actually, John Wren, if you look at Omnicom Advertising Group before they acquired IPG, had a better strategy. He was trying to push them together. Publicis didn’t nudge. They were quite forceful. And it was painful for them. They were more violent than probably I would have been. But directionally, it was the same thing.

 

What do the holdcos look like in three years?

 

I don’t know what [WPP CEO] Cindy [Rose] is going to do. She has huge balance sheet problems. But I would guess Omnicom will still be here. Publicis will still be here. Dentsu will be in a different form. Havas will be in a different form. WPP will be in a different form.

 

Let’s talk about S4. You guys had a tech winter. Where are you in all of this?

 

It’s tough. We rely on wholesale AI adoption. We really depend on more car verticals, more financial services verticals, more packaged goods. If we can get a little bit more traction, it will be fine. But we need more traction. And we’re not organized enough. Our integration has not been perfect by any means. Liquidity is much better.

 

What does that traction look like and how do you get it?

 

Traction looks like building relationships. If you look at our biggest relationships—Google, Amazon, T-Mobile, Disney, GM, Meta—there’s good stuff there, but the problem is that 45% of it is tech. And they’re all spending money on capex.

Sunday, March 29, 2026

17419: Sorrell Subscribes To Payment Scheme.

 

Digiday reported S4 Capital CEO Sir Martin Sorrell is seeking to move clients from billable hours to subscriptions.

 

In recent years, White holding companies and White advertising agencies have been generating the most creative work in accounting departments, finding innovative ways to cook the books.

 

So, it’s no surprise Sorrell—who has always been a financial schemer—would pioneer in payment ploys.

 

Being a pioneer, however, has not yet made Sorrell a profiteer. He might soon concede to work for peanuts.

 

“It’s about change management”: Sir Martin Sorrell says the billable hour is dying, but getting clients to move in is proving harder

 

By Seb Joseph

 

Yes, outcome-based pay is en vogue. But none of it matters if agencies aren’t using AI at the scale the economics require. And that only happens if clients are willing to pay for it. Most aren’t.

 

S4 Capital CEO Sir Martin Sorrell said as much on his marketing group’s latest earnings call yesterday — a cold dose of realism about the hype that comes after the billable hour. For S4, the answer is subscriptions: fixed annual fees bundling senior talent with agentic workflows, brand-specific knowledge bases and quarterly software-style upgrades. 

 

Sceptics call it the agency retainer rebranded. Believers see it as an on-ramp to a world beyond the billable hour. The truth is probably somewhere in between — at least for now. The reasons are outlined [elsewhere], but Sorrell’s own framing cuts to it. 

“The far more important thing is, what is going to be the pace of AI adoption?,” he told analysts on the call. “Because the simple fact is, consumers have adopted AI faster than companies, and companies aren’t doing it because it’s not just about technology or workflow — it’s about change management. And companies find it difficult to do that.”

 

The ones that do are buying into subscriptions. One enterprise client signed up last year and restructured its entire agency relationship around the model. Three more are in discussions with more expected to follow since subscriptions are now baked into every new business pitch S4 runs. By year-end, it wants a quarter of its revenue running that way.

 

Getting there is another matter. For subscriptions to work at scale, a lot has to align: existing clients have to be willing to reopen contracts; procurement teams have to get comfortable pricing outputs rather than hours; rising inference costs have to stay absorbable within a fixed fee; and the pace of AI adoption across the broader market has to quicken. None of those things are moving as fast as S4 would like.

 

“I don’t think anybody in marketing procurement teams is against a change,” said Wes ter Haar, co-founder of Media.Monks and executive director at S4Capital on the call. “Everybody understands that the traditional model really isn’t sustainable in any meaningful way, but doing this at scale in a running business is not the easiest thing to do.”

 

There are, however, pockets of momentum. Chiefly, in automotive and financial services — sectors pushed by existential threat rather than enthusiasm. Chinese EV competition in one case, fintech disruption in the other. FMCG, arguably the most valuable category for marketing groups, is starting to stir too. And counterintuitively, the conflict in the Middle East could accelerate things further: if its reverberations push inflation higher and keep rates elevated, that macro pressure may yet do what enthusiasm alone hasn’t: force the kind of AI adoption at scale that makes the new model viable.

 

“It may be that that acts as a catalyst,” Sorrell said on the call. “If global growth slows, inflation rises and interest rates are stickier — which seems to be the scenario already being built in by some of the analysts and the investment banking firms — that might be the engine for increased tech adoption and AI adoption.”

 

Should that happen, it might finally give the agency model’s critics pause. So far, nothing has — not the workforce automation, the M&A, nor the push into agentic services.

 

Or as as Gartner vp analyst Jay Wilson, put it: “More advanced client-side organizations are starting to pivot to enterprise-wide platforms from the hyperscalers — and the agency AI platforms, which are basically just connectors of Salesforce and Adobe and Workfront, we don’t believe are going to be as relevant going forward.”

 

Which is why the subscription model’s internal economics matter as much as its commercial logic. The pitch to clients is simple: as AI improves, they get more for the same price. Fifty assets a month becomes seventy, no fee increase. But that only works if S4 can keep absorbing rising compute costs inside the fixed fee. With video models and always-on usage expanding fast, pass-through costs may eventually be unavoidable. That would mean variable pricing — the one thing procurement teams are least equipped to sign off on.

 

“On personalization at scale, I would say there’s more opportunity,” said Sorrell. “But on visualization and copywriting, we’re seeing compression — if you charge on time, that is compressed.”

 

As it stands, it’s the latter two where most AI work done at scale is happening. CMOs, despite what they say, see AI more as an efficiency play than an effectiveness one — technology to get things done faster and cheaper, not necessarily better. AI only compounds for a marketing services company like S4 if it can do all three. Without the effectiveness part — the bit where real outcomes can be pegged to results — the agency model becomes an even slipperier slope to commoditization.

 

How S4 got to this point is a sobering illustration. The group reported revenue of £754.8m for 2025, down 11% on a year earlier, with net revenue falling 10.8% to £673m. Nearly half that revenue historically came from big tech — but marketing spend at Amazon, Meta and Alphabet has been essentially flat since 2022, while their AI infrastructure spending has grown over 133%.

 

Tech clients spending more on AI and less on agencies: that’s the burning platform behind S4’s model shift, and the reason subscriptions are no longer just a commercial experiment.

Monday, August 25, 2025

17164: Short News On S4 Capital & MSQ Partners.

 

More About Advertising reported the proposed acquisition deal between S4 and MSQ is DOA, which is good news for MSQ, as Sir Martin Sorrell surely would’ve added FU to the acronym blender—prompting the marketing group to consider changing their masthead to MSQ QQ.

 

Plus, group shots like the slightly doctored MSQ portrait depicted below might’ve required rethinking and reframing—unless Sorrell felt comfortable standing on a few crates.

 

MSQ and S4 Capital deal is off – for now anyway

 

By Stephen Foster

 

Oh well, it was fun while it lasted (briefly.) Marketing group MSQ has firmly ruled out a merger with Sir Martin Sorrell’s S4 Capital, saying that such discussions may have taken place between representatives of its majority shareholder One Equity Partners (OEP) and S4 Capital, but its board was not involved.

 

It added that neither OEP nor MSQ (below) intends to pursue further discussions regarding the transaction. Which does rather suggest that OEP and MSQ haven’t exactly been on the same page.

 

In truth it was hard to see the logic behind such a deal, at least from MSQ’s point of view. S4 is much bigger, at least in terms of numbers, but struggling as its tech clients do their own thing with AI. MSQ, on the face of it although it’s hard to know with P/E-owned businesses, seems to be doing OK on its own although it’s currently embarked on expansion in the US, which has derailed many a UK-based marcoms business.

 

Sorrell also has the equivalent of a golden share in S4 which means, in essence, it can’t do anything major without his say-so and there are no signs he’s ready to leave the stage just yet. Which may be a relief to nervous MSQ employees.

 

As with many UK businesses the problem for both S4 and MSQ is scaling. You may grow by acquiring lots of smaller companies but, some time or other, there’s going to be one that doesn’t work and that can bring the whole edifice down. Quoted Next 15 may rue the day it bought what was left of Engine Group to become House337 just when the market for such agencies turned down (it also had a run at M&C Saatchi.)

 

Back in the day the brothers’ Saatchi & Saatchi bought Ted Bates in the US for a then eye-watering $450m (boss Bob Jacoby trousered $110m only to be fired later) beginning the rapid decline which led to its eventual sale to Publicis. S4 Capital was once valued at a dizzy £4.5 billion (it’s now £136m/$184m.) Sorrell is reported to have rebuffed a $700m approach from Mark Penn’s Stagwell last year.

 

It is, as they say, complicated. There’s surely another twist or two in the S4 Capital tale. MSQ may be better off charting its own course but, as some of the above details illustrate, it’s far from easy.

Tuesday, August 12, 2025

17152: S4 + MSQ = OMG WTF.

Advertising Age republished a Bloomberg News report revealing Corporate Colonizer Sir Martin Sorrell is in merger talks, whereby his S4 Capital would colonize, er, acquire MSQ Partners.

 

Sorrell is quick to critique his original monstrosity WPP—as well as the holding company model he developed—yet doesn’t hesitate to replay the acquisition aggressor with his latest mad man experiment.

 

Hey, maybe incoming WPP CEO Cindy Rose should ignite her strategic review by staging a hostile takeover of S4 Capital.

 

Martin Sorrell’s S4 says it’s in talks to combine with MSQ

 

S4 Capital Plc, the advertising agency founded by Martin Sorrell, is in talks to combine with private equity-owned MSQ Partners.

 

Any deal would be structured so that S4 would be taking over MSQ, though talks are at a preliminary stage and may not lead to a deal, S4 said in a statement on Monday. MSQ is owned by private equity firm One Equity Partners. S4 was responding to an earlier report about the potential combination on Sky News.

 

Sorrell founded S4 in 2018 after he left WPP Plc. The company has struggled at times to compete with its larger rivals and most recently warned that clients were being more cautious with spending ahead of new tariffs from the U.S. Shares have declined about 35% this year giving S4 a market value of about £131.7 million ($177 million) when the market closed on Friday.

 

The company had attracted takeover offers from ad firm Stagwell last year, the Wall Street Journal had reported at the time. Sorrell rebuffed the offers as too low, though his reluctance to consider bids has frustrated some of the firm’s senior officers, the Journal said, citing people familiar with the matter who it didn’t name.

 

—Bloomberg News

Monday, July 28, 2025

17138: Sir Martin’s Musing, Muttering, Mumbling, Madness.

 

Digiday published more pointed pissing from Sir Martin Sorrell, who appears to be confusing thought leadership with resentment; that is, the old man believes he’s delivering insight, yet he’s just spewing insolence.

 

As Digiday notes, Sorrell is quick to critique the White holding company business model he refined and developed to spectacular disaster. However, his purported solution—S4—has not proven a viable alternative. In fact, S4 might wind up being just another failed venture.

 

Of course, Sorrell never acknowledges the human costs associated with his mad experiments. How many people have lost their livelihoods because of ill-conceived mergers, blundering takeovers, and inept management?

 

Such indifference separates leaders from losers. And louts.

 

As S4 struggles, Sir Martin Sorrell keeps firing shots at the industry he created

 

By Seb Joseph and Michael BĂ¼rgi

 

Sir Martin Sorrell has never lacked for sharp opinions. That’s expected from the architect of the holdco company era. But lately, his commentary has gone from reflective to biting.

 

In recent months, he dismissed WPP’s consolidation efforts as a “disgrace”, lobbed shade at Omnicom for what he implied would be an overpriced acquisition of IPG and blamed WPP’s current struggles on what he called “weak leadership” under CEO Mark Read. In June, he went further: WPP, he said, might be beyond saving. 

 

Needless to say, Sir Martin makes his views known. 

 

“People, including you, ask for my views,” said Sir Martin in an email to Digiday in response to a request for comment on this story. “I’ve spent almost 50 years at Saatchi & Saatchi, WPP and S4Capital — perhaps, as a continued shareholder in WPP, I have some helpful experience and views.”

 

Helpful or not, those views come with baggage. S4 Capital, the company he built in response to the holdco model he now criticizes, has lost 97% of its market value, issued multiple profit warnings and been caught in the crosswinds of shifting ad budgets. His pointed commentary hasn’t gone unnoticed.

 

“I only wish he was as vocal and clear minded on what needs to be done at S4,” said Anthony Freedman, CEO of marketing services group Common Interest and a shareholder of S4 Capital, on LinkedIn.

 

Sir Martin sees no contradiction. He’s clear-eyed about S4’s challenges — but sees them as macroeconomic not structural. The plan, as he told Digiday, is to stay the course — tighten operations, ride out the volatility and focus on growth in the Americas, Middle East and APAC where clients are still chasing gains. In Europe, he’s betting on a push for greater efficiency and effectiveness.

 

S4 Capital, he insisted, was built for moments like this — when clients demand more for less and faster. 

 

The company is focused “on top-line growth with our existing clients and new ones, improved operating margins and liquidity, focussing particularly on improved pricing, billability and reducing duplication,” he wrote. 

 

But markets don’t trade on messaging alone. In September 2021, S4 Capital’s shares peaked at $12.25. Today, they hover near 30 cents. 

 

Whether there’s an actual bounceback is unclear — but Sir Martin sees one. He points to a current blue-chip client list that includes Alphabet, Meta, Amazon, GM, T-Mobile and Walmart as well as a newly won CPG advertiser he declined to name, as proof that the model still works. As for the broader slowdown in ad dollars from tech clients? That was inevitable, he said. 

 

“With 50% of our almost $1B of revenues coming from tech there has been pressure on opex and hence marketing as the “Magnificient 7” and others spend over half a trillion dollars on AI-related capex,” wrote Sir Martin. 

 

What’s certain is this: Sir Martin’s critique of the holding company model may still resonate, but the industry feels that it hits differently when coming from someone whose own reinvention is under just as much strain. 

 

“What you have to question is, does he provide enough day-to-day leadership?” said one source who knows Sir Martin well.

 

But it wasn’t always like this. 

 

In its early days, S4 Capital had momentum. Launched in 2018 with Sir Martin freshly out of WPP, the company moved quickly — acquiring MediaMonks and MightyHive within months, and landing marquee clients like Procter & Gamble, Nestle, Mondelez and Bayer in its first year. 

 

“I always thought the work from them was outstanding, and a big part of that was down to the people who worked there,” said a senior marketer familiar with the group’s pitch, speaking on condition of anonymity. “They weren’t just typical ad agency people, they were from tech and consulting backgrounds.”

 

For a time, this was the story of S4 Capital: proof that a tech-first approach could disrupt the industry. But as the years went on, the cracks began to show.

 

Audit delays in 2022 rocked investor confidence. The following year brought revised revenue guidance — twice — along with 500 layoffs and pressure on margins. The stock cratered. By 2024, S4 Capital was still playing defense, issuing fresh profit warnings and watching its tech-heavy client base pull back on spending. In the background, a CFO change signaled yet another attempt to restore stability. 

 

“The person at the top is all about finance models and therefore doesn’t put anywhere near enough emphasis on people and the actual operating of the business,” said an exec familiar with S4’s plans, who exchanged anonymity for candor.

 

As the issues mounted up, so did the tensions at the top. S4 Capital’s equity-heavy acquisition model, which had once helped close deals, began to backfire as the share price fell. Founders who had bought into Sir Martin’s vision started to disengage. 

 

“The model of 50% cash and 50% equity works when the share price is growing,” the same exec said. “But as soon as it starts falling, all the owners of the acquired companies become extremely disgruntled, demotivated and less productive.”

 

Despite these issues, few in the industry are ready to count Sir Martin out. His track record, stubbornness and proximity to power still matter. But there’s a growing sense — even amongst those who admire him — that the public markets may no longer be the right arena. 

 

“In 12 months’ time, I believe S4 Capital will not exist in its current form,” said an ad exec, with knowledge of the internal issues at the business. 

 

Industry chatter suggests that delisting is one option. A merger or acquisition is another. Some believe parts of the business could be spun out through a management buyout. However it unfolds, the status quo isn’t expected to hold. 

 

Sir Martin did not comment on any of these potential routes for the company.

 

“I don’t think any public company chairman could or should answer that,” he wrote.” At S4 Capital, I’m focussed on long-term shareowner value maximization and acting in the best interests of all shareowners.”

 

Where S4 will end up is anyone’s guess. It certainly can’t keep headed in the same direction it’s been going. Sir Martin knows what’s wrong with holding companies, but has not yet proven he knows how to build the alternative.

Tuesday, October 29, 2024

16822: Chatting About Bias In AI.

 

Adweek published a perspective on addressing bias in AI. Written by an experiential creative and copywriter at Monk, the exposition presented challenges that emerged during the invention of AI-powered robot Sir Martian, named after Sir Martin Sorrell. Um, it sounds like AI bias inspired by a biased A-hole.

 

 

4 Ways to Mitigate Bias in AI and Close the Diversity Deficit

 

Key lessons from a 2024 Cannes activation

 

By Larissa Pontez

 

Feed a prompt to an AI image generator and you’re bound to encounter an insidious pattern: Do the people look … too stunning? Perhaps even wanton? 

 

Gender, race, body type, nationality, religion—you’re almost guaranteed to get prejudiced and outdated stereotypes when using these descriptors in prompts. And “wanton” is a deliberate adjective; it’s mostly used pejoratively toward women, and AI tends to oversexualize female images. These glaring imbalances showcase a recurring problem with AI outputs: the replication of societal biases, which can be harmful to actual people and communities. 

 

I wrestled with this firsthand while helping develop Sir Martian, one of our key AI demos featured at Cannes earlier this year. Sir Martian, playfully named after Sir Martin Sorrell, is an AI-powered robot in the form of an alien caricaturist. Throughout the festival, he invited attendees to sit down for a quick chat and a sketched portrait, based on their appearance and tastes. 

 

I’m proud that the demo was a success, because as you can imagine, this interaction was more than a simple conversation. And it taught me a lot about the privileges and responsibilities of shaping a new technology. Here’s what I learned. 

 

Words matter—your data sets the tone 

 

Most AI tools available for the general public are trained on datasets that aren’t accessible or visible to users, so I feel particularly fortunate to work at a company that creates and trains its own models. It really is a “great power, great responsibility” scenario. 

 

The foundation of any generative AI model should be diverse and comprehensive. By expanding the range of base images and training materials, developers can create AI systems that represent a broader spectrum of human experiences. This enriches outputs and helps combat entrenched biases. 

 

With Sir Martian, specificity was essential for aligning user inputs with desired outputs. After some trial and error, we found that we had to train the model combining visual input with very precise text prompts in order to get it to represent people accurately. 

When given a picture of a Black woman and the prompt “woman with braids,” the AI model automatically defaulted to a woman with German-style braids. We had to train and fine-tune it using specific terms like “cornrows” and “box braids” to get it to create accurate drawings. Giving the system a wider variety of terms to connect to visual references was crucial to getting more diverse depictions. 

 

This step was humbling because I encountered my own limitations in the process. For example, we don’t have a large Muslim population where I’m based in Brazil, yet a global audience traveling to Cannes would likely include women in hijabs or chadors. This prompted me to research the nuances between different articles of dress that, to an untrained eye, may have been seen as interchangeable. The experience highlighted the importance of stepping outside of our bubbles to recognize what we don’t know, in order to learn and incorporate diverse cultural elements that better serve global users. 

 

Diversity is (and isn’t) everyone’s responsibility 

 

As the only woman on the team building Sir Martian, the problematic depiction of women raised alarm bells for me early on but didn’t faze my male colleagues until I brought it to their attention. We need more diverse teams who can authentically lead AI in the right direction. But at the same time, the onus shouldn’t be on minorities alone to fix biases that have affected them for generations. 

 

Overcoming these biases demands collective effort. After I discovered flaws in Sir Martian’s AI model, I partnered closely with a developer on the project who was dedicated to addressing these issues. I reached out to a Black co-worker and Muslim women in our global community for their feedback on whether Sir Martian’s drawings were respectfully reflecting their identities. These are just some examples of the cross-disciplinary collaboration that needs to happen in order to make a change; once you flip the switch and understand what needs to be done, the rate of progress is astounding. 

 

The industry has a ways to go, but we’re seeing positive change. Since Sir Martian launched, we’ve instated a global AI policy to help staff become more conscious of common biases that occur in AI systems, such as data bias, algorithmic bias, and confirmation bias. Perhaps more importantly, fostering an inclusive environment encourages a shared responsibility in creating AI systems that accurately and fairly reflect diverse experiences, ultimately benefiting everyone. 

 

Know where to draw the line, and back up your decisions 

 

Our industry celebrates how AI will unlock personalization for everyone, but there are limits. The unfortunate reality is that, when it comes to accurately depicting everyone, we can’t perfectly address every difference on every project. But we can try to be as thorough as possible given the limits of technology, time, and budgets. 

 

When it comes to being more diverse and inclusive, for example, people naturally focus on accounting for a variety of skin tones. That’s great, but it’s often as far as we go. What about different body types and sizes? How might a generated portrait differ when someone is sitting in a wheelchair instead of standing up?

 

We should not only address these questions, but also begin asking them at a project’s inception. Those of us developing consumer-facing generative AI activations must be conscious of where our parameters fall, as well as able to justify the decisions we make. 

 

When working on Sir Martian for the demo in Cannes, we decided to leave children out of the training data, knowing that they were not our target audience. This was a conscious decision rather than a blind spot in our process, as representation and inclusion so often are in AI projects.

 

It’s time to do better 

 

We all know that AI is an amazing tool that has progressed by leaps and bounds over the last few years, but one thing it can’t do is correct our own blind spots. That’s on us to identify and address. 

 

AI serves as a mirror to our society, reflecting both its progress and its persistent challenges. If left unchecked, biases can become even more ingrained through AI. Tackling this issue isn’t a task for minorities alone—it’s something we all need to work on together. This shared commitment can help genuinely turn AI into a force for positive change.