The Big Read Technology sector

How Microsoft spread its bets

beyond OpenAI

The tech giant has worked to execute an AI

strategy independent of Sam Altman’s start-up
following a leadership crisis last year

Madhumita Murgia in London and Stephen Morris in San Francisco YESTERDAY

In late November last year, as India faced Australia in the final of the Cricket
World Cup, mega-fan Satya Nadella was distracted. He was dealing with a
work crisis.

Nadella, who runs $3tn software giant Microsoft, had learned just days
earlier that Sam Altman, the chief executive of OpenAI, the start-up in which
Microsoft has invested $13bn, had been fired by his board in a surprise coup
for not being “consistently candid”.

Caught unawares despite being OpenAI’s largest financial backer, Nadella

moved quickly to fix the disruption. Once reassured Altman had not done
anything egregious, he pushed first to hire and later reinstate the
entrepreneur, in an attempt to restore stability at the start-up with which
Microsoft’s future was now closely threaded.
In all, it took Microsoft’s leadership 10 days of intense work to repair the
fallout from the aborted coup.

For Microsoft and its investors, the incident was a reminder of how central
OpenAI had become to its strategy: the growth of artificial intelligence.
Nadella’s decision to bet on the start-up in July 2019, long before its flagship
product ChatGPT became a household name, had created one of the tech
industry’s most successful partnerships.

You are seeing a snapshot of an interactive graphic. This is most likely due to
being offline or JavaScript being disabled in your browser.
Not only did it give the software company a head start in the booming
market for generative AI, but Microsoft’s share price has more than tripled
since the initial $1bn investment five years ago, allowing it to compete with
Apple for the title of world’s most valuable company and widen its
advantage over arch-rival Google. Speaking in a Financial Times interview
early last year, Nadella said Microsoft and OpenAI had developed a “mutual
dependence”.

But in the eight months since the board dispute, the tech giant has worked to
execute an AI strategy independent of Altman’s start-up. It has diversified its
investments and partnerships in generative AI, built its own smaller,
cheaper models, and hired aggressively to develop its consumer AI efforts.

In February, Microsoft announced a multiyear partnership and investment

into French AI start-up Mistral; the following month it paid another peer
Inflection — led by Google DeepMind co-founder Mustafa Suleyman —
$650mn to license its technology and hire most of its talent; and then in April
invested $1.5bn in Abu Dhabi AI group G42.

That same month, it also announced it had built its own family of generative
AI models known as Phi-3 — software that is smaller in size and complexity,
and cheaper to run than so-called large language models such as OpenAI’s
GPT-4. Microsoft has said its Phi-3 models are being used by the likes of
BlackRock and Epic, and have outperformed GPT-3.5, an earlier version of
OpenAI’s model, which ran its chatbot ChatGPT.

As the company’s vast spending on AI

continues — accounting for much of its
As one of the most $56bn in annual capex — investors and
valuable companies in regulators are closely scrutinising the
the world . . . you can’t high-profile alliance with OpenAI, and
have your eggs in one Microsoft’s strategy to challenge Google
basket on its home turf: search.
“Before November, I didn’t think they had a diversification strategy. Satya is
one of the smartest executives and leaders you can ever find in the
ecosystem. If after the experience in November he is not thinking about
diversification, I would be worried,” says Navrina Singh, chief executive of
Credo AI, who worked on commercialising AI systems at Microsoft until
2019. “As one of the most valuable companies in the world . . . you can’t have
your eggs in one basket. You can’t be blinded by innovation.”

Microsoft’s efforts to expand its AI ecosystem have changed the terms of its
relationship with OpenAI, and also exposed the flaws within it. “I think you
can see some fractures of trust and once those fractures appear it’s very
difficult to reduce or remove them,” Singh adds.

A sales executive at Microsoft says it is just smart business. “The other

partnerships are a safeguard, not just if OpenAI goes down but in case a new
start-up comes up with something better,” the person says. “What happens if
Mistral, Cohere or Microsoft bring out a better model, what does Sam have?
Huge consumer reach, good researchers, but if the best model isn’t GPT4
then who cares?”

Since its leadership crisis, OpenAI has replaced its board almost entirely,
although its governance structures remain largely unchanged.

Altman was reinstated as a director in March, following an independent

review conducted by a law firm into the events, which concluded that his
behaviour “did not mandate removal”. In the aftermath, Microsoft was first
given, and then withdrew from, an observer seat on the board, amid
growing scrutiny by antitrust regulators.

But in recent months, OpenAI has been rocked by internal rows and high-
profile resignations. This week, the company’s president, former board
member and prominent co-founder Greg Brockman announced a leave of
absence until the end of the year - in order, he later said, to spend time with
his family. Brockman was one of Altman’s fiercest supporters during the
November coup, when he resigned from his role in protest, before rejoining
days later. At the time, Nadella offered him a job at Microsoft, alongside
Altman.
 You are seeing a snapshot of an interactive graphic. This is most likely due to
being offline or JavaScript being disabled in your browser.

In May, former chief scientist and co-founder Ilya Sutskever quit to found his
own AI company, after playing a leading role in the failed attempt to oust
Altman, for reasons he never elaborated on. The raft of departures mean
that nine of the start-up’s 11 co-founders are currently not working there.

Another recent exit, Jan Leike, who led OpenAI’s efforts to steer and control
super-powerful AI tools and worked closely with Sutskever, said his
differences with the company leadership had “reached a breaking point” as
“safety culture and processes have taken a back seat to shiny products”.

He and others have gone to work for rival Anthropic, which itself was
founded by former OpenAI employees who broke with Altman and the rest
of OpenAI’s leadership in 2021.
According to former Microsoft employees, this is not the first time OpenAI
has operated in a dysfunctional manner. Sophia Velastegui, former chief AI
technology officer for business applications at Microsoft, says that even prior
to ChatGPT, some of the product launches had not been communicated to
Microsoft as expected. “OpenAI still operates like a start-up in many ways, so
their tolerance for risk is higher than Microsoft’s.”

Altman continues to have powerful supporters in Silicon Valley. LinkedIn co-

founder and Microsoft board member Reid Hoffman describes Altman as a
“hall of fame entrepreneur” who does not suffer from the same “messiah
complex” as some other prominent founders.

Still, recent departures and changes at OpenAI will leave the tech giant’s
leadership more nervous about management maturity at the start-up, and
provide a timely reminder that Microsoft cannot be overly dependent on any
one third-party technology in the AI vertical.

“Aligning expectations about how and

when to communicate is a process when
OpenAI still operates a disrupter like OpenAI joins forces with
like a start-up in many an established player like Microsoft,”
ways, so their tolerance says Velastegui. “At the end of the day,
for risk is higher than both companies are still learning how
Microsoft’s best to work together.”

While investments in G42 and Mistral

were not necessarily knee jerk
responses to Altman’s ouster, those deals took on more significance as a way
of reassuring nervous investors that the tech giant was spreading its bets.

More controversially, the so-called “acqui-hire” of Inflection founder

Suleyman and most of the start-up’s staff in March set Microsoft on a path to
confrontation with its biggest AI partner. The combative former Google
DeepMind executive, who left that company having developed a reputation
as a bully, was put in charge of a new internal AI unit at Microsoft and
tasked with building consumer-facing products that would compete with
those from Altman’s OpenAI.

According to multiple people in the tech industry, there are already tensions
simmering between the ambitious pair.
There will be more complications down the line. The US Federal Trade
Commission is probing whether the Inflection deal was structured to
circumvent antitrust laws, essentially gutting the smaller company of talent
and software, while avoiding the formal scrutiny a full takeover would have
brought. The FTC has also opened an investigation into the OpenAI
partnership, resulting in Microsoft proactively dropping its board observer
seat.

Despite the scrutiny, the Inflection deal has become a model for other tech
giants seeking talent. In June, Amazon hired most of the staff at AI-agent
start-up Adept and paid $330mn to license its intellectual property. Last
week, Google rehired the founder of chatbot maker Character.ai and paid
more than $2bn to license its technology and cash out existing investors.

The rash of buyouts underlines the trend of power flowing away from the
start-ups like OpenAI, which kick-started the AI revolution, back to Big Tech
gatekeepers, cementing the hold they’ve had on the sector for decades.

“[OpenAI] remains a strong partner and we are pretty confident they have
solved their internal issues,” says Eric Boyd, corporate vice-president of
Microsoft’s Azure AI cloud computing platform, who manages the
relationship with OpenAI. “At least to me, there has not been a particular
strategic shift as a result of what happened.”

Brad Lightcap, OpenAI’s chief operating officer, says: “While we have

evolved from a small start-up to a company serving the world’s largest
companies, Microsoft remains an important partner.” Its funds and
infrastructure have helped “enable OpenAI to innovate and deliver
groundbreaking research and products,” he adds.

But as Altman’s vaulting ambitions grow — from plans to build trillion-

dollar Middle Eastern-financed chip factories to AI-centric smartphones with
Japan’s SoftBank — the two companies find themselves increasingly in
competition.
In June, Apple said it would integrate ChatGPT into its operating systems,
giving the start-up access to its 2.2bn active devices around the world.
Notably, ChatGPT has not been integrated into Windows in the same
fashion.

OpenAI is hiring rapidly for a sales team to pitch their products to

commercial clients directly, going after the companies that Microsoft wants
for its Azure platform with the same underlying technology that powers its
workplace AI assistant, Copilot.

Boyd insisted that although the two companies collaborated on creating

models, “we go to market and approach customers completely
independently . . . If customers ask us what the difference is in the offerings,
we tend to point to the ways that we show up as a company — OpenAI is a
start-up and we’ve been around for decades.”

He suggests that, as a start-up, OpenAI has fewer checks and balances than
its established partner. “We have a long history of working with enterprises,
handling sensitive data . . . We know how to do privacy and compliance.”
 You are seeing a snapshot of an interactive graphic. This is most likely due to
being offline or JavaScript being disabled in your browser.

Ultimately, though, even if Microsoft loses a pitch to OpenAI, it still wins —

although the reverse is not true. Azure is OpenAI’s exclusive cloud provider
and will be paid for whatever computing power it uses, Boyd says. Microsoft
is also agnostic about which AI models are used, so long as they are accessed
through its cloud.

“We have over 1,600 models available through Azure AI . . . the main thing
we want is people to be building and using them on Azure,” he says.

Microsoft has been keen to play up the burgeoning rivalry with its partner in
light of escalating antitrust scrutiny. In its 2024 annual report, OpenAI was
added to its list of direct competitors in AI, search and advertising. It also
flagged that it has “limited ability to control or influence third parties with
whom we have arrangements, which may impact our ability to realise the
anticipated benefits”.
The difference in strategy between Microsoft and Google is stark. The search
giant is attempting to build a “full stack” of AI in-house, from LLMs and
consumer-facing chatbots to hardware such as chips and servers in its cloud
business.

The deal with OpenAI means that “Microsoft has decided to outsource their
AI R&D,” says one Google executive, who asked to remain anonymous. “We
are being more cautious.”

He compares the current moment in AI to a scene in Shakespeare’s play

Macbeth when a character asks a trio of witches to “look into the seeds of
time” to determine which will grow. “AI feels like asking those witches [to
predict the future]. We’ve seen 100,000 seeds planted and we don’t yet know
which will grow.”

Investors are starting to question the heavy spending on AI by Big Tech,

which reached a combined $106bn in the first six months of 2024. After a
historic bull run, the tech-dominated Nasdaq has fallen 13 per cent from its
mid-July record peak, helping spark a wider market rout.

Microsoft reported that capex had jumped 80 per cent in the fourth quarter
and it had spent $56bn in its financial year 2024 — about half on
infrastructure such as data centres and land, with the remainder on chips
and server capacity. Ben Reitzes, an analyst at Melius Research, says
executives’ comments “imply an aggregate figure of at least $80bn for 2025”.

Some of this spending is driving the ambitions of OpenAI: “We have also
increased our investments in the development and deployment of
specialised supercomputing systems to accelerate OpenAI’s research,”
Microsoft said in its annual report.

Still, analysts were impressed by early tangible evidence of a translation of

investment into earnings. Chief financial officer Amy Hood predicted a
strong ramp up in AI-related profits in the second half of next year and
Nadella said Azure AI now had 60,000 customers, up more than 60 per cent
from a year ago.
“Microsoft continues to be the clear beneficiary from Generative AI
initiatives, with 46 per cent of chief investment officers citing Microsoft as
gaining the largest share of IT spending over the next one and three years,”
says Morgan Stanley analyst Keith Weiss, referring to a survey the
investment bank conducted. “The number two vendor, Amazon, was cited by
just 6 per cent.”

Even as the OpenAI drama was ongoing, Nadella cast himself as the
dominant partner in the relationship.

“We were very confident in our own ability. If tomorrow OpenAI

disappeared, I don’t want any customer of ours to be worried about it,” he
said in a November interview. “We have all of the [IP] rights to continue the
innovation . . . We have the people, we have the compute, we have the data,
we have everything.”

Additional reporting by George Hammond and Camilla Hodgson

Copyright The Financial Times Limited 2024. All rights reserved.