It has been over three-and-a-half years since the debut of OpenAI’s ChatGPT spurred booming corporate interest in generative artificial intelligence tools to remake work. And yet Hitachi, which employs nearly 290,000 globally, still hasn’t deployed a single, enterprise-wide AI tool for all workers across the Japanese conglomerate.
This more cautious approach may prove to be prudent for Hitachi, given that research has shown a high number of enterprise AI pilots fail and a debate that’s intensified regarding the cost of AI, leading most large employers like Hitachi to closely track workplace AI usage as costs rise.
Bala Krishnapillai, the senior vice president and chief information officer of Hitachi’s Americas division, says there are plenty of AI tools that have been widely embraced by the company’s workforce. His internal AI adoption strategy focuses on three buckets. The first are everyday productivity tools like Microsoft Copilot and Google Gemini, used to summarize emails, meetings notes, and for translation, the latter especially critical for the 607 subsidiaries that operate across 190 global markets.
Krishnapillai says his IT team works closely with business leaders to evaluate and approve job-specific tools, which can include AI-enabled content creation used by creative professionals or for competitive analysis to help the sales team. The third focus area is on developers and AI coding assistants, where Hitachi works closely with Anthropic.
“From the enterprise AI strategy standpoint, there is no one solution,” says Krishnapillai, who joined Hitachi in 2018 and has served as CIO of the Americas division since April 2025.
Though he encourages AI adoption across the company, Krishnapillai says “consuming of tokens is a big thing. We are putting in some controls.” Some departments where competitiveness is deemed critical to their work, like research and development, have no AI usage restrictions. But more broadly across Hitachi, division managers are responsible for how their teams use AI and are tracking their spending.
One bigger challenge that Krishnapillai has had to address is data complexity. Founded in 1910 as a single mining machinery repair shop, Hitachi today operates a sprawling business that includes rail systems, digital products, industrial machinery, power and renewable energy, and medical systems. Ranked #197 on the Fortune Global 500, Hitachi was most well known by the general public for manufacturing and selling consumer electronics including televisions and camcorders, a business it has fully exited as it pivoted to selling more digital systems and services, which now account for 27% of revenue.
Hitachi generates $70 billion in annual revenue today and it got that massive through acquisitions. Some of the company’s larger deals in recent years include the $11 billion spent on the power grids business acquired from Swiss-based tech firm ABB and a $9.6 billion deal to scoop up U.S. software vendor GlobalLogic.
A century of dealmaking has resulted in data stored across more than 150 different customer relationship management (CRM) software systems, including Salesforce, SAP, and Microsoft. When marketing and sales teams would work on a new business proposal, it could take weeks to produce an accurate analysis report.
“We have massive data stored in our ecosystem,” says Krishnapillai. “It was unmanageable from an IT enterprise standpoint.”
Krishnapillai tapped enterprise software vendor Appian to connect data on top of the legacy infrastructure, without requiring any migration to a single database. The two main benefits to this approach is that today, teams are able to create new projects at both a faster pace and with sharper insights from Hitachi’s data ecosystem.
“When we create a proposal now, it becomes stronger, more compelling, and very competitive,” says Krishnapillai.
Hitachi and Appian say the new approach has led to a 40% efficiency gain for the sales and marketing team, as well as a 20% reduction in operating costs.
Matt Calkins, CEO of Appian, says this data fabric will also make it easier for Hitachi to embrace agentic AI. He says these autonomous agents can, at times, ask unexpected questions and search for unanticipated data sources across the business in order to reach their conclusions.
“Everybody’s got scattered data, and everybody needs to inform unpredictable agents,” says Calkins. “They’ll be like humans, exploring the enterprise and making decisions, and so they need to go places that you didn’t already orchestrate and expect.”
That’s an optimistic view, if everything goes right. Rogue AI agents—a report published this week said models built by Anthropic and OpenAI took unsanctioned actions—highlight the dangers of embracing AI in this manner and the need for tight governance.
Krishnapillai says he’s focused on establishing security protocols and data protection and privacy before the company is fully ready to embrace autonomous tasks. He says Hitachi is actively talking with vendors to find a software system that can monitor the creation of all AI agents, which will help avoid unnecessary duplication, but also promote agents that can be applied in cross-functional ways across the business.
“We are in the early phase,” says Krishnapillai, regarding his progress on agentic AI. “Our goal is to become autonomous in the future. But we are not there yet.”
John Kell
NEWS PACKETS
**White House, AI firms meet to discuss model testing. **On Tuesday, the Trump administration hosted representatives from several AI companies— Fortune reported that Anthropic, OpenAI, Google, and Microsoft were all involved—to discuss a potential new framework that will review the cybersecurity capabilities of new models before they are released publicly. The discussion comes
several weeksafter Trump signed an executive order in June that established a framework for the U.S. government to vet the national security risks of these AI systems for up to a month before they were widely released. AI developers were allowed to voluntarily opt into the process. There are quite a few subplots lurking in the background: advancements from China present a looming competitive threat to the U.S. AI hyperscalers,
OpenAI’s Hugging Face hackraised alarms about the industry’s risks, and not all AI firms are on the same page about whether tighter regulations will support healthy competition.
**Investors cheer Alibaba’s new AI model. **Alibaba, meanwhile, earned investor praise on Monday after it released its “most powerful” AI model called Qwen3.8-Max. The Chinese tech behemoth said that the new AI model’s capabilities include coding, research, and visual intelligence and that it delivered comparable, and in some cases better scores than Anthropic’s Fable 5, according to CNBC. Fable 5 debuted in June but was offline for a few weeks after the Trump administration said it presented a national security concern before returning online by early July after export control restrictions were lifted following close coordination between the U.S. government and Anthropic.
**Apple-OpenAI lawsuit gets testy. **Litigation between the makers of the iPhone and ChatGPT intensified this week, with Apple filing a motion in federal court seeking a preliminary injunction to stop OpenAI from using what it called stolen trade secrets, according to Bloomberg. The outlet reports that Apple’s lawyers allege that if its trade secrets were “spread throughout OpenAI, or get embedded in its products and operations, the damage cannot be undone.” OpenAI disputed the legal request and said “we do not have, nor want, any of their trade secrets.” The legal squabble began in July when Apple initially sued OpenAI over trade secrets and alleged that OpenAI had hired more than 400 former Apple employees. The years-long relationship between the pair has been turbulent for months, as Apple previously leveraged OpenAI’s tech to infuse AI into its Siri digital assistant, but later switched to Google’s Gemini.
**AI’s talent war is still hot. **As top talent continuously shuffles between large AI hyperscalers like Anthropic and OpenAI, more established tech giants including Google and Meta, and AI labs, Axios reports that loyalty is hard to command even with sky-high compensation packages. One big, notable shuffle happened at Thinking Machines, where co-founder Lilian Weng announced she would step down due to health concerns, only to immediately rejoin OpenAI. The Wall Street Journal reports that the talent war isn’t just playing out among Silicon Valley companies. Wall Street firms are also vying for a very small pool of math college graduates and Ph.D.s who have been able to command salaries typically within the $350,000 to $500,000 range, though there have been instances where pay for a prized entry-level employee has exceeded $1 million.
ADOPTION CURVE
**After years of pushing AI adoption, high costs are propelling a pivot. **Since the debut of ChatGPT in late 2022, corporations and their C-suites have aggressively pushed employees to embrace the new technology. Coding tools, assistants, and agents have proliferated, as have AI training courses and hackathons to promote greater usage. But a new survey from consulting giant EY finds that leaders may be rethinking this narrative, and it’s because the AI bill is coming due.
An astonishing 98% of those surveyed who have invested in AI and use tools that require AI tokens say such usage and related costs “have caused their organization to reconsider their approach.” And yet, only 64% of the senior leaders surveyed said their enterprise is actively monitoring AI token usage and has established clear budget guidelines for how much it can spend.
The EY US AI Pulse survey found that 37% of senior leaders are reconsidering how their companies evaluate AI vendor pricing models, while 35% are reevaluating how they budget for AI and 32% are looking at who gets access to these tools. These figures indicate that the short-lived era of AI tool proliferation for all employees across an entire company may already be ending.
Dan Diasio, EY global consulting AI leader, tells Fortune that organizations are “putting a price on intelligence” and starting to look at which employees need to have access to the latest and greatest frontier versus open-source models, and whether there should be limits on how much AI they can access.
“That does not mean that companies are pulling back or slowing down,” says Diasio, of AI usage within the workplace. “But, they're accelerating with a mindset towards controls and deciding who gets access to what sort of technology to be able to do the work the best way possible.”
Courtesy of EY
JOBS RADAR
Hiring:
- **Moneycorp **is [seeking a head of technology](https://www.linkedin.com/jobs/view/4447571861/?alternateChannel=search&eBP=CwEAAAGfyk0AY020ieYcvOjPkaL6wvGnLc2POWZmDHOPmraMsOh8U5tpN0rxtsA4OMA-pFaj208mrXPnl8cvMM1r0sW1eGbaE9kre25Oe8xsdU1WgFSfr8UbXjX2pFpDMKZWIQzmJWDDwujxJj0mdgSYdS2sHfr7Q1ucIW5aF4r2FbQ1_WwRP-ZPyD6hRnTg2tQNCKne4NR2FMOHB-NftFrY0z0p8SoHD1fa7Or7t27lAHj4nCmWfZFBC1bIc2MtWG044eMGSaXsR64io9CsQUlnUPHV4yIxwHq2x_TQxb8JJLyzEuFQR_rdJRZJttYtwbCSvPZ3t-IxDmJeSs1c1p_XIF2TYdY9OFe16PST7Vp5NS081mTUQlmKxuNcBbz-5yRoL0VOS0c7Xgro9-0WnoE4cc_j1Ec4opTuzADTErmpJo5-16UujWQqAGNqFVS1HuCiCpjkJPR4ZlmyXpgwvy6X-C-1KZNlWUnW3KEJhte6Kc9T8Q&refId=GrUO%2BX%2FlaJQlt8pUxvM5qg%3D%3D&trackingId=L0R7QvIJZL10Noz3hx1CTw%3D%3D), based in Stamford, Connecticut. Posted salary range: $300K-$350K/year.
- **Cape** is [seeking a chief information security officer](https://www.linkedin.com/jobs/view/4446935623/?alternateChannel=search&eBP=CwEAAAGfymNQfZlmLzs21XpDe1u5gIHciT--TpMVQxFN0rmoZh-EUYPbK2Snaha1cHh2H1jw1iVzyNM0mI6BgezOfBmxXwvOAmMyvqvxinJAdMK90r_gTmzqLyboYSGy-c305u7MK5Q-QgRHDsMPNmxFSXgZW2I5agaBUqk0K0y_kLms0Nty97lB3UG3QZtWeoJkA1i6fk0LkDtyui6QYjmJEQNi7jxASZ54isx9huqEXGnC-0IcO9d_xV3ab5QnaINDgbzHwhMzt6QiWpa6-rt-hmtHtim9AzN6Q_sRz3jRFS4mJvxdIe_VuPoW6ebWJu_eCwfElhhm1JuUqqrycJPaI6FhwEdVKUBCZcFynms9ORaNi5oACxPQ9NTL157BhJgXGfPnmr_HtFQ3Vt50S5kQf48fEHITfwvtriipR7duRLG_GfiUfw5ZF8C-FtfMvUdOFh7HZWwRxTTUhu783ZbNcgfjqVzFk2U5NNoJHdFCkw_esw&refId=AUCyvtt3Y1eaW%2FUh8PLAsQ%3D%3D&trackingId=dKM8LwdqcVMUkwEeqQ71xw%3D%3D), based in New York or Washington. Posted salary range: $300K-$350K/year.
- **Adonis **is [seeking a head of AI](https://www.linkedin.com/jobs/view/4408930828/?alternateChannel=search&eBP=CwEAAAGfymTpAgKouEpxX3yA7PLjUqsBtXN7FR12wiaHTvnoqL1mG7Xs4x-kGV_3esO_4Sfv1RyPeihGuqpC_Lz1K1V-l655CpI2ZvD1wiqrzNs7FEt02pmXIB6EDzHDvQ0EN08jnAtFUHYOSMS3TPVxyae3WDrMa0WKRcqkn5-2OVDMA2s05jv9y53FWE8GezJQpOkMKh40hCldhx1_A5oY1idBdUdLqY2jYZoYBdg9PZ6KwwDAKT3xdaV8IoV0ZDwL50ivzJDE7zvklwhAd3YO32A5VrbYLKVlCXUsFKa5CDSaW-czVqS1pJFX0y0Az5V9JyKfXIA0WUQcJ-ckkrHZpCKUPputitWqZikxCAFeCU2q7EEpkYBBqsZ1VFOfbiRN3h0rRaO10r7IZDsfNbge-PZ_MbugzjiqmRz6x2AyNI0fbq8ID_Fq-9V22v1a4-3bgrZG3MnvvW9BkMiRiUQAiZ0TInFmUOIx5gCBCGXPjqgGuhee&refId=X8YmixayxpXzxoE%2FjITjdA%3D%3D&trackingId=5cmaxahDlgQioaPmX9FZGQ%3D%3D), based in New York. Posted salary range: $250K-$275K/year.
- Estée Lauder is seeking a VP of value chain data and AI, based in New York. Posted salary range: $211.6K-$377.2K/year.
Hired:
**- Coinbase **promoted Rob Witoff as CTO, after initially joining the crypto exchange in November 2024 as head of platform. Prior to joining Coinbase, Witoff was the co-founder and CEO of crypto infrastructure and security provider Unit 410. He also previously served as CTO at Polychain Capital and worked at Coinbase from 2014 to 2017, ending his first tenure as chief architect.
**- Freshworks **named Ryan Manning to serve as chief product and technology officer, reporting to CEO Dennis Woodside. He joins the software development company after most recently serving as chief product officer at software vendor BMC Helix. Manning previously served as a director at ServiceNow and a manager at Dell.
**- Keller Logistics **appointed Mel Hudson-Nowak as CTO, joining the logistics provider after most recently serving as VP of IT compliance and risk management at GM Financial, the financial services division of auto giant General Motors. Previously, Hudson-Nowak served as VP of technology at US Foods, where she spent more than 12 years in various leadership roles.
**- **Crexendo promoted Chris Aaker to the role of CTO, effective August 1. Most recently, Aaker served as VP of engineering at the business communications software provider. Aaker initially joined NetSapiens, which Crexendo acquired in 2021, nearly two decades ago and has earned several promotions throughout his tenure.
**- ****Camden National Bank **promoted Josh Nash to serve as CTO, succeeding William "Bill" Martel, who is retiring from the New England regional bank. In his new role, Nash will lead enterprise technology strategy, which includes infrastructure, cybersecurity, and data governance. Nash initially joined the bank in 2008 as an assistant vice president.
**- ****Qu **appointed Paul Rubin as chief product and technology officer, where he will lead the restaurant technology company’s product and engineering roadmap. Rubin joins Qu from the foodservice technology firm ParTech, where worked for 12 years, most recently as chief strategy officer.
**- **Diversified announced the appointment of Pete Emminger as CTO, joining the tech and media firm to lead its go-to-market strategy. Most recently, Emminger served as CTO at media services vendor NEP Group. He also previously served as a VP at Blizzard Entertainment.
Sign upfor free.