Verified
Intel
Piloting- Semiconductors
- 85,100 employees
- iGPT and 1AI, built in-house
- 30,000+ monthly iGPT users (Jan 2025)
- $305M estimated AI value (2024)
Intel deploys AI inside the company through iGPT and 1AI, platforms its IT organization built that run open models on Intel's own Xeon and Gaudi hardware. More than 30,000 employees a month used iGPT, a no-code assistant builder, in Jan 2025, and 1AI, an agent platform, was first deployed for sales and marketing. Intel IT estimated the business value of Intel's AI use cases at $305 million for 2024, without a published method. In Jul 2026 Intel began adding Google's Gemini Enterprise across engineering, supply chain and corporate operations.
intel.comTheir approach to AI
- The platform
- Intel IT built iForge, the back end that connects models and data sources, and iGPT, a no-code front end on which employees had built more than 7,000 personal assistants by Jan 2025. 1AI, an agent platform first deployed in Sales and Marketing, served more than 2,000 users in Oct 2025.
- The models
- Intel runs open models such as Llama 3 70B, Qwen3 and Mistral 7B on its own Xeon and Gaudi hardware and sends complex, low-volume reasoning to frontier cloud models. Intel added Google's Gemini Enterprise in Jul 2026 and built its Ask Intel support assistant on Microsoft Copilot Studio.
- The reach
- More than 30,000 Intel employees used iGPT each month as of Jan 1, 2025, out of about 100,000 at the time. Intel had 85,100 employees at Dec 27, 2025 and has published no newer usage count.
- The catch
- Intel has reported no AI return in cost, revenue or headcount. Intel's $305 million figure for 2024 is Intel IT's own estimate of business value, blending productivity, cost and revenue with no published method. Intel IT has reported no newer figure since Cindy Stoddard became CIO in Dec 2025.
Snapshot
- Employees
- 85,100
- Dec 27, 2025, after the 2025 Restructuring Plan (2025 Form 10-K)
- Technology budget
- 1.6% of revenue (2024)
- IT spending of $8,240 per employee; Intel IT reports spending as a share of revenue
- AI/ML headcount
- Not disclosed
- Intel IT had about 2,607 employees in Jan 2025
- Who owns AI
- Cindy Stoddard, CIO
- Since Dec 1, 2025, reporting to CEO Lip-Bu Tan, who also leads the AI product groups
- Platform
- iGPT on iForge, and 1AI
- Built by Intel IT; Gemini Enterprise added from Jul 2026
- Models
- Open models, frontier models, Gemini
- Llama, Qwen and Mistral on Xeon and Gaudi (Feb 2026)
- Reach
- 30,000+ monthly iGPT users
- Jan 2025, of about 100,000 employees then
- Use cases
- 436 tracked, 274 in production
- Intel IT AI dashboard, Jan 2025
Key findings
Intel IT reported in Jun 2025 that the AI use cases on its company-wide dashboard produced an estimated $305 million of business value in 2024 and projected $376 million for 2025. Intel IT counted productivity, cost reduction, cost mitigation and incremental revenue, without publishing how the value was measured.
More than 30,000 Intel employees used iGPT, Intel IT's no-code generative AI platform, each month as of Jan 1, 2025, and had built more than 7,000 personal assistants on it. Intel has published no newer usage count.
Intel runs most of its internal AI on its own chips to lower cost per token (Feb 2026). iGPT and the 1AI agent platform use open models such as Llama 3 70B, Qwen3 and Mistral 7B on Xeon processors and Gaudi accelerators, on-premises and in IBM Cloud and Denvr Dataworks.
Intel said on Jul 16, 2026 that it will deploy Google's Gemini Enterprise across engineering, supply chain and corporate operations, moving beyond isolated AI pilots. Intel will also use Google Cloud capacity for chip design simulations.
Intel's AI-driven data center workload agents recovered more than 0.5 million CPU cores and removed the need for new capital purchases through 2025, Intel IT reported in Jun 2026, without putting a dollar value on the capital avoided.
Where Intel sits
- Stage
- Piloting
Intel has reported no business return that meets the ROI rule, for any use case or for the company as a whole. Intel IT's 2024-25 Annual Performance Report (Jun 2025) estimated $305 million of business value from AI use cases in 2024. The $305 million figure blends productivity, cost reduction, cost mitigation and incremental revenue with no published method, and Intel's figures for facilities, IT support and data centers are hours saved, workload, capacity or forecasts. Intel's rollout record puts it at Piloting. The same report counted more than 30,000 monthly users of iGPT, the company-wide generative AI platform Intel IT runs, as of Jan 1, 2025. CIO Cindy Stoddard has owned Intel IT and its internal AI-enabled transformation since Dec 1, 2025, reporting to CEO Lip-Bu Tan. That rollout evidence leaves the 18-month window in Dec 2026 unless Intel publishes a newer count. Intel falls short of Scaling because it has published no AI-credited return in cost saved, revenue, headcount or output per employee in any function, and no firm-level AI figure measured in those terms.
How the stages are defined - Ownership
- CIO Cindy Stoddard runs Intel IT, reporting to CEO Lip-Bu Tan
- Platform
- Hybrid: iGPT and 1AI built in-house; Gemini Enterprise bought
- Models
- Open models on Intel hardware, frontier cloud models, Gemini
- Hosting
- Intel's own data centers and private cloud on Xeon and Gaudi, IBM Cloud, Denvr Dataworks, Google Cloud, Microsoft Copilot Studio, and AI PCs
- Reach
- 30,000+ monthly iGPT users (Jan 2025)
How they deploy
Build the platform, run it on Intel silicon
Intel IT built its own generative AI platform, iForge as the back end and iGPT as the no-code front end, and runs its inference on Intel Xeon processors and Gaudi accelerators. Intel IT said in Feb 2026 that it expects a lower cost per token from its own infrastructure than from public cloud providers. Intel's factories were moving GPU-based deep learning training to Gaudi accelerators (Jun 2025).
One agent platform instead of siloed chatbots
Intel IT built 1AI to consolidate separate chatbots into one interface. 1AI's semantic layer routes each prompt to the most relevant agent and to a model chosen by relevance and cost, from commercial frontier models to open-source and fine-tuned ones. Intel IT and the Sales and Marketing group cataloged more than 60 use cases in early 2025. The 2 teams picked the first by strategic fit, users affected, hours saved or revenue impact, data maturity and total cost of ownership (Oct 2025).
Small models where they suffice
Intel IT wrote in Feb 2026 that not every use case needs a frontier model. To control per-token costs, Intel IT uses models such as Llama 3 70B and Qwen3 30B, or small models such as Mistral 7B and Qwen3 8B. Intel IT's infrastructure agent platform runs high-volume inference on open models on Xeon processors on-premises and keeps frontier cloud models for complex, low-volume reasoning (Sep 2026).
More bought tools under Lip-Bu Tan
Intel moved large parts of its marketing work to Accenture, which uses AI for information processing, task automation and personalized communications (Jun 2025). Intel built its Ask Intel customer support assistant on Microsoft Copilot Studio (Feb 2026) and chose Google's Gemini Enterprise as a central hub for employees to build and deploy agents (Jul 2026). Lip-Bu Tan said in Nov 2025 that the CIO role matters more as Intel builds deeper partnerships with AI platform companies.
Use cases approved by value
Intel IT tracks every AI use case across Intel on one dashboard. Intel IT approves new use cases on cost of implementation, redundancy with existing solutions, usage and value, with the Information Security and Privacy teams in the approval process (Jun 2025). Intel IT funnels generative AI use cases onto its enterprise platform so that they use approved, internally built capabilities.
- Deployment model
- Intel runs internal AI as central platforms built by Intel IT with use cases set by the business units. Intel IT has been led by CIO Cindy Stoddard since Dec 1, 2025. Intel IT runs iGPT for all employees, 1AI for business unit agents, the AI dashboard that tracks every use case, and agent platforms for its own infrastructure and service desk. Business units build on these platforms or alongside them. Intel's Sales and Marketing group has its own AI Strategy team, which chose the first 1AI use cases. Intel Foundry's Industrial Automation and Facilities teams run their own agentic AI for fab utilities (Jan 2026). Since Jul 2026 Intel has added Gemini Enterprise as a hub where business functions build line-of-business agents.
- Model providers
- Intel's 1AI platform lists OpenAI, Inflection AI and IBM Watson among the commercial models it can route to, alongside open-source and fine-tuned models (Oct 2025). Intel IT runs Llama 3 70B, Qwen3 30B, Mistral 7B and Qwen3 8B behind iGPT through an OpenAI-compatible stack with a LiteLLM gateway and Langfuse monitoring (Feb 2026). Intel IT's infrastructure agents pair open models on Xeon with unnamed frontier cloud models (Sep 2026). Intel added Google's Gemini models through Gemini Enterprise in Jul 2026, and Intel's customer-facing Ask Intel assistant is built on Microsoft Copilot Studio (Feb 2026).
- Data and infrastructure
- Intel IT operates 55 data center modules at 15 sites with 160 megawatts of capacity and more than 457,000 servers (Apr 2026). Intel IT runs iGPT's general-purpose inference on-premises and in the Denvr Dataworks cloud, and uses IBM Cloud for Sales and Marketing AI use cases, on Xeon processors and Gaudi accelerators (Feb 2026). Intel hosts 1AI in its private cloud (Oct 2025). Intel's chip designers run more than 308 million compute-intensive batch jobs a week, and from Jul 2026 Google Cloud C4 and N4 instances add capacity to Intel's on-premises design compute. Intel IT had deployed 25,000 AI PCs by Jan 2025 so that some AI workloads run on the laptop.
Platforms and internal tools
iGPT and iForgePilot by May 2024
iGPT is Intel IT's no-code front end on which any Intel employee can build and use AI assistants, and iForge is the back end that runs and connects models and data sources. Intel IT calls iGPT its Citizen AI platform; Intel's AskIT IT support bot moved from a commercial product onto iGPT in 2025.
- Users
- 30,000+ monthly users and 7,000+ personal assistants (Jan 2025)
- Models
- Llama 3 70B, Qwen3, Mistral 7B and other open models
1AI (One AI)2025
1AI is Intel IT's agentic AI platform, built on an open-source front end with custom agents hosted in Intel's private cloud, that routes each prompt to a specialized agent and model. Intel showed an early version at Intel Vision 2025. The first deployment, in Sales and Marketing, covers a CRM chatbot, seller self-support, a support agent chatbot, marketing claims guidance and co-marketing help.
- Users
- 2,000+ (Oct 2025)
- Models
- OpenAI, Inflection AI, IBM Watson, open-source and fine-tuned models
Gemini EnterpriseJul 2026
Gemini Enterprise is Google's agent platform, which Intel is integrating across its workforce in engineering, supply chain and corporate operations, with agentic coding assistance and line-of-business agents. Early pilots include agents that recommend subject matter experts and draft executive messaging.
- Models
- Google Gemini
Ask IntelFeb 2026
Ask Intel is a generative AI support assistant on Intel's support site, built on Microsoft Copilot Studio, that opens cases, checks warranty coverage, gives troubleshooting guidance and hands off to human agents. Intel launched Ask Intel after removing inbound support phone numbers in most countries.
Facilities agentic AIBy Jan 2026
Intel's Industrial Automation and Facility Operations teams run a multi-agent system over fab utilities such as ultrapure water and chilled water. The system uses LangGraph orchestration, Model Context Protocol and living digital twins, and its agents draft work orders that a human must approve.
IaaS agentic AI platformBy Sep 2026
Intel IT's infrastructure agent platform, in production across its hosting environment, combines live telemetry with configuration, incident and change data so that agents can answer plain-language questions and suggest root causes. The platform uses open models on Xeon for high-volume inference and frontier cloud models for complex reasoning.
AI data center workload agentsBy 2025
Intel IT runs machine-learning agents in its design data centers that predict resource needs, place jobs, reallocate capacity, detect idle jobs and end problematic ones.
Vendors and partners
Google CloudJul 2026 expansion
Gemini Enterprise for employees; C4 and N4 capacity for chip design simulations · Intel said on Jul 16, 2026 that the expansion builds on a multi-year strategic collaboration, and Google Cloud augments, not replaces, Intel's on-premises compute.
IBM CloudBy Feb 2026
Hosts Intel Sales and Marketing group AI use cases on Intel hardware
Denvr DataworksBy Feb 2026
Cloud platform behind part of iGPT's general-purpose inference
MicrosoftFeb 2026
Copilot Studio, the platform behind the Ask Intel support assistant
AccentureJun 2025
Runs large parts of Intel's marketing with AI tools · Intel did not say how many marketing roles were cut; Intel called Accenture a longtime partner (Tom's Hardware, Jun 2025).
OpenAI, Inflection AI and IBM WatsonBy Oct 2025
Commercial models 1AI can route prompts to
FormantBy Oct 2026
Industrial AI partner in a Foundry Facilities alarm-triage proof of concept
Deloitte and Gartner2025
Assessed Intel IT's data, analytics and AI maturity
Leadership
AI, data and platform leaders
Executives
AI org map
Central IT-built platforms, use cases owned by business units
Cindy Stoddard, appointed Intel's CIO in Nov 2025, runs Intel IT and its internal AI platforms and reports directly to CEO Lip-Bu Tan. Intel IT builds iGPT, 1AI and the dashboard that tracks every AI use case. Business units such as the Sales and Marketing group and Intel Foundry set their own use cases, and some run their own AI teams. Lip-Bu Tan has also led Intel's AI and Advanced Technologies groups directly since Sachin Katti left in Nov 2025, but those groups build products for customers, not internal tools.
Intel's internal AI work sits in Intel IT and in business teams that build on its platforms. Intel IT had about 2,607 employees in Jan 2025 and builds iGPT, 1AI and the AI dashboard. The business teams include the Sales and Marketing group's AI Strategy team and Intel Foundry's Industrial Automation and Facilities engineers. Intel has not published how many AI or machine learning specialists it employs for internal use. Intel's AI product groups, which Lip-Bu Tan has led directly since Nov 2025, build the chips and software Intel sells.
Reporting lines
Chief Executive Officer
Intel IT
Cindy Stoddard, SVP and CIO
Builds and runs iGPT, 1AI, the AI dashboard and IT agent platforms
AI and Advanced Technologies groups
Product groups led by the CEO since Nov 2025; not the internal AI program
Sales and Marketing Group
Leader and reporting line not confirmed from a primary source
SMG AI Strategy
Boaz Efroni Rotman, Director of AI Strategy and Roadmap
Intel Foundry
Industrial Automation and Facility Operations
Run agentic AI over fab utilities with Intel IT
Intel's Nov 18, 2025 release says Cindy Stoddard reports directly to Lip-Bu Tan. Intel has not published an organization chart for its internal AI work; the Sales and Marketing and Foundry units on this chart come from IT@Intel white paper author titles.
Where the AI workforce sits
Intel IT
Estimate: about 2,607 IT employees against about 100,000 employees in Jan 2025; Intel does not count its internal AI specialists separately.
~3%Business units and Intel Foundry
Estimate: the remainder of the workforce, where the iGPT users, the Sales and Marketing AI Strategy team and the Foundry automation engineers sit.
~97%
Named people
Use cases
Intel tracked 436 AI use cases on Intel IT's dashboard in Jan 2025, 274 of them in production, across manufacturing, chip design, IT, sales and marketing and support. Intel has published no return that meets the ROI rule for a single named use case: Intel's facilities, IT support, data center and design figures are hours saved, ticket or workload reductions, recovered capacity or forward estimates. For AI as a whole, Intel IT estimated in Jun 2025 that the use cases on its dashboard delivered $305 million of business value in 2024. The $305 million figure blends productivity, cost and revenue with no published method, rather than measuring cost saved or revenue. Intel IT had credited AI with more than $1.4 billion of value in 2023 under a measure it described separately.
| Use case | Area | Status | ROI |
|---|---|---|---|
| Fab utilities monitoring and maintenance Intel's facilities agents predict ultrapure water resin exhaustion 90 days ahead instead of 30, detect facility anomalies 2 to 5 days earlier, tune chilled-water plants with digital twins and draft work orders for human approval. Intel reports 30% faster anomaly detection, 15% to 20% less technician workload and more than 120,000 labor hours saved a year. Intel estimates about $7.5 million of annual savings from the ultrapure water monitoring (Jan 2026). | Manufacturing | In production | |
| Yield and process analysis Intel uses AI and machine learning, together with digital twins of facilities, equipment, process and logistics, for design-technology co-optimization, advanced process control and yield analysis on advanced nodes such as Intel 18A. An Intel 18A business line manager described the work in a talk. | Manufacturing | In production | |
| Factory equipment and quality control Intel engineers deploy AI across hundreds of manufacturing use cases, including closed-loop control of manufacturing equipment, computer vision quality control and preventive maintenance (Mar 2025). Intel's factories are moving deep learning model training from GPUs to Gaudi accelerators (Jun 2025). | Manufacturing | In production | |
| Alarm triage for Foundry remote operations centers Intel Foundry Facilities and Formant tested agents that give remote operations engineers context on industrial waste system alarms; the proof of concept saved 581 technician hours over 540 days (Oct 2026). | Manufacturing | Pilot | |
| Design data center workload agents Intel IT's machine-learning agents place design jobs, predict resource needs and end problematic jobs in data centers where Intel's design engineers run more than 308 million batch jobs a week. Intel IT says the agents recovered more than 0.5 million CPU cores and removed the need for new capital purchases through 2025, without a dollar figure (Jun 2026). | Chip design | In production | |
| Agentic coding and design workflows on Google Cloud Intel is adding Gemini-based agentic coding assistance and engineering automation, and is scaling chip design simulations onto Google Cloud C4 and N4 instances to run more of them at once (Jul 2026). | Chip design | Announced | |
| 1AI agents for sellers Intel's Sales and Marketing group uses 1AI agents for CRM questions, seller self-support, support knowledge search, marketing performance claims guidance and co-marketing help. The agents serve more than 2,000 users with answers in seconds instead of hours (Oct 2025). | Sales and marketing | In production | |
| Marketing work moved to Accenture and AI Intel handed large parts of its marketing work to Accenture, which uses AI for information processing, task automation and personalized communications, leaving lean in-house teams (Jun 2025). | Sales and marketing | In production | |
| Communications agents Intel is piloting Google Cloud agents that recommend subject matter experts, draft executive-ready messaging and create supporting materials across communications channels (Jul 2026). | Sales and marketing | Pilot | |
| Ask Intel support assistant Ask Intel, built on Microsoft Copilot Studio, opens support cases, checks warranty coverage and escalates to human agents for customers and partners. An Intel spokesperson reported better satisfaction and case resolution rates, without figures (Feb 2026). | Customer support | In production | |
| AskIT employee support bot Intel moved its AskIT IT support bot from a commercial product onto iGPT. By mid-May 2025 more than 11,000 employees had held more than 26,000 conversations with AskIT, and Intel IT reported 30% fewer "How do I" tickets, with a goal of 50% by the end of 2025. | IT | In production | |
| Infrastructure operations agents Intel IT's infrastructure agent platform answers questions across infrastructure domains, enriches incident analysis and compresses tasks that took hours into minutes, reclaiming more than 120 engineering hours a month (Sep 2026). | IT | In production | |
| IT incident ticket clustering Intel IT built an unsupervised clustering framework on Xeon with open-source models that sorts its IT incident tickets into a hierarchy of categories so staff can find known fixes faster (Jul 2026). | IT | In production | |
| Personal assistants on iGPT Intel employees build and share no-code AI assistants on iGPT; Intel IT counted more than 7,000 personal assistants by Jan 2025 and is adding agents for legal contract reviews and "How do I" questions. | Firmwide | In production |
Adoption
Intel spread AI in layers. Machine learning has run in manufacturing, yield analysis and sales for years; Intel IT's generative AI platform followed from a 2024 pilot, then agents in 2025 and 2026. iGPT began as a pilot whose engaged users almost doubled in 30 days (May 2024). By Jan 1, 2025, when Intel had about 100,000 employees, iGPT had more than 30,000 monthly users and more than 7,000 personal assistants. Intel IT had estimated in 2024 that about 75% of employees would eventually use iGPT. The 1AI platform served more than 2,000 users in Oct 2025, and Intel IT had rolled out 25,000 AI PCs by Jan 2025. Intel has published no usage count since Jan 2025, through the restructuring that left 85,100 employees at Dec 27, 2025 and the Gemini Enterprise rollout announced in Jul 2026.
By the numbers
Rollout, first pilot to today
2018
Intel IT developed Sales AI, a platform that let Intel scale its sales activity, according to Intel IT's Oct 2025 account.
Mar 2024
Intel's chief people officer described AI Inside, the program that focuses Intel's internal AI on manufacturing, software, chip design and chip validation, and iGPT as the in-house model employees are encouraged to use.
Apr 2024
Intel IT reported that Intel's AI efforts contributed more than $1.4 billion in value in 2023, measured across all AI solutions in production.
May 2024
Intel IT described its generative AI platform as a pilot whose engaged users almost doubled in 30 days, and estimated that about 75% of Intel employees would eventually use it.
Mar 2025
Intel IT reported that engineers were deploying AI across hundreds of manufacturing use cases and that it had built a new generative AI platform for teams across Intel.
Jun 2025
Intel IT reported more than 30,000 monthly iGPT users, 436 tracked AI use cases with 274 in production, 25,000 AI PCs and an estimated $305 million of AI business value in 2024.
Jun 2025
Intel told marketing staff that much of their work would move to Accenture, which uses AI to handle tasks Intel staff did.
Oct 2025
Intel IT reported that its 1AI agent platform was in use in Sales and Marketing with more than 2,000 users and would scale to other business units.
Feb 2026
Intel IT said it runs iGPT inference on Xeon and Gaudi on-premises and in the Denvr Dataworks cloud, and Sales and Marketing AI in IBM Cloud.
Jul 2026
Intel announced it will deploy Gemini Enterprise across engineering, supply chain and corporate operations, moving beyond isolated AI pilots.
Sep 2026
Intel IT reported that its infrastructure agent platform is in production across its hosting environment.
Economics & controls
Spend and ROI
Technology budget
Intel IT reports its spending as a share of revenue rather than as a dollar budget. Intel's IT spending was 1.6% of revenue and $8,240 per employee in 2024, down from 2.1% and $8,900 in 2023 (Jun 2025). Intel spent $13.8 billion on research and development in 2025, down 17% from 2024 mainly because of headcount reductions (2025 Form 10-K). Intel IT runs 55 data center modules at 15 sites with 160 megawatts and more than 457,000 servers (Apr 2026), and says its data center strategy saved $13.36 billion from 2010 to 2025.
What AI itself costs
Intel has not disclosed what it spends on AI tools or tokens for its own staff. Intel IT runs iGPT on open models on Xeon and Gaudi to lower cost per token and chose 1AI's models partly by cost (Feb 2026, Oct 2025). Intel IT's infrastructure agents send high-volume inference to on-premises open models to cut token costs, keeping frontier cloud models for complex reasoning (Sep 2026).
Reported ROI
Intel IT reported in Jun 2025, in an annual report signed by then CIO Motti Finkelstein, that the AI use cases on its dashboard delivered an estimated $305 million of business value and 321,680 hours saved in 2024. Intel IT projected $376 million and 696,030 hours for 2025, counting productivity, cost reduction, cost mitigation and incremental revenue. Intel has not published how it measures the value or whether hours are converted into dollars. Intel IT's previous report (Apr 2024) said Intel's AI efforts contributed more than $1.4 billion in value in 2023, measured across all AI solutions in production. The 2024 report counted revenue from product improvements or sales, or cost savings in development and manufacturing. Intel has not explained the gap between the 2 figures. Intel has published no AI value figure for 2025 or 2026, and none of Intel's figures is stated as cost saved in currency, revenue, headcount or output per employee.
Productivity markers
Intel's other AI results measure time, workload or capacity, not cost. AskIT cut "How do I" IT tickets by 30% (May 2025), and facilities agents saved more than 120,000 labor hours a year and cut technician workload 15% to 20% (Jan 2026). Intel IT's infrastructure agents reclaim more than 120 engineering hours a month (Sep 2026). Intel's design data center agents recovered more than 0.5 million CPU cores and removed the need for new capital purchases through 2025 (Jun 2026). Intel IT estimates that ultrapure water monitoring will save about $7.5 million a year (Jan 2026), a forecast. Intel IT credits its data center strategy, automation, and AI and machine learning together with supporting 3,563% more volume over 16 years with 3% higher headcount costs.
Headcount
Intel had 85,100 employees at Dec 27, 2025, including Mobileye and other subsidiaries, after a 2025 Restructuring Plan that cut management layers and reduced the core workforce by about 15% by the end of fiscal 2025, compared with its Q2 2025 headcount. Intel has not credited those cuts to AI. In Jun 2025 Intel did move large parts of its marketing work to Accenture and its AI tools, without saying how many roles were affected. Intel IT's Aug 2026 digital employee experience paper describes an enterprise of more than 120,000 employees, a figure that does not match the 10-K count.
Governance and risk
Use-case approval
Intel IT approves AI use cases on cost of implementation, redundancy, usage and value, with the Information Security and Privacy teams in the process (Jun 2025). Intel IT tracks every use case on one AI dashboard. Intel's 2024 governance process routes generative AI applications through a model and use-case approval owned by Information Security and Privacy. The process applies standards from Intel's ethical AI team, and projects on the platform inherit its security controls.
Data boundary
Intel IT hosts 1AI in Intel's private cloud (Oct 2025) and keeps high-volume infrastructure inference on on-premises open models to retain control of sensitive operational data (Sep 2026). Intel's 2025 Form 10-K warns that wider use of generative AI models within its internal systems, processes and tools may create new attack methods for threat actors.
Humans approve agent actions
Intel's facilities agents draft work orders automatically but need human approval before any action runs. Intel IT says operator trust grew through step-by-step automation in which AI proposes and humans approve (Jan 2026). Intel IT's infrastructure agents use human-approved write actions (Sep 2026).
People decisions
Intel's chief people officer said in Mar 2024 that Intel does not use AI to make hiring, internal job movement or promotion decisions, does not use it to write performance reviews and does not monitor employee email with it.
Customer-facing AI
Intel's Ask Intel assistant warns that the accuracy of its answers cannot be guaranteed. Intel retains chat logs, which Intel and third-party providers may process under Intel's privacy policy (Tom's Hardware, Feb 2026). Intel's 2025 Form 10-K says Intel may be at a competitive disadvantage if it does not adopt AI in its operations as effectively as competitors.
Recent signals
Oct 28, 2026
Intel Foundry Facilities and Formant published a proof of concept in which agents triaged industrial waste system alarms for remote operations centers, saving 581 technician hours over 540 days.
Sep 8, 2026
Intel IT reported that its infrastructure agent platform runs in production, pairing open models on Xeon with frontier cloud models and reclaiming more than 120 engineering hours a month.
Jul 16, 2026
Intel and Google Cloud said Intel will deploy Gemini Enterprise across its workforce and use Google Cloud C4 and N4 instances for chip design simulations. CIO Cindy Stoddard called Gemini Enterprise a central hub for employees to build and deploy agents.
Jun 17, 2026
Intel's 2025-26 Corporate Responsibility Report said Intel launched a Manager AI chat series and learning platforms to build AI proficiency among managers.
Jun 2026
Intel IT reported that AI-driven workload agents in its design data centers recovered more than 0.5 million CPU cores and removed the need for new capital purchases through 2025.
Feb 26, 2026
Intel IT said it runs iGPT on Llama, Qwen and Mistral models on Xeon and Gaudi, on-premises and in the Denvr Dataworks cloud, with Sales and Marketing AI in IBM Cloud.
Feb 22, 2026
Intel launched Ask Intel, a support assistant built on Microsoft Copilot Studio, after removing inbound support phone numbers in most countries.
Jan 23, 2026
Intel's 2025 Form 10-K reported 85,100 employees and warned that wider use of generative AI in its internal systems may open new attack methods.
Jan 2026
Intel IT reported that agentic AI in Intel's fab utilities saves more than 120,000 labor hours a year and cuts technician workload 15% to 20%.
Dec 1, 2025
Cindy Stoddard took office as Intel's CIO, reporting directly to Lip-Bu Tan, with a remit that includes Intel's internal AI-enabled IT transformation.
Nov 2025
Sachin Katti, Intel's chief technology and AI officer since Apr 2025, left for OpenAI, and Lip-Bu Tan took over the AI and Advanced Technologies groups.
Quick answers
- How does Intel deploy AI?
- Intel deploys AI inside the company through iGPT and 1AI, platforms its IT organization built that run open models on Intel's own Xeon and Gaudi hardware. More than 30,000 employees a month used iGPT, a no-code assistant builder, in Jan 2025, and 1AI, an agent platform, was first deployed for sales and marketing. Intel IT estimated the business value of Intel's AI use cases at $305 million for 2024, without a published method. In Jul 2026 Intel began adding Google's Gemini Enterprise across engineering, supply chain and corporate operations.
- What does Intel use AI for?
- Intel tracked 436 AI use cases on Intel IT's dashboard in Jan 2025, 274 of them in production, across manufacturing, chip design, IT, sales and marketing and support. Intel has published no return that meets the ROI rule for a single named use case: Intel's facilities, IT support, data center and design figures are hours saved, ticket or workload reductions, recovered capacity or forward estimates. For AI as a whole, Intel IT estimated in Jun 2025 that the use cases on its dashboard delivered $305 million of business value in 2024. The $305 million figure blends productivity, cost and revenue with no published method, rather than measuring cost saved or revenue. Intel IT had credited AI with more than $1.4 billion of value in 2023 under a measure it described separately.
- Who leads AI at Intel?
- AI at Intel is led by Cindy Stoddard (Senior Vice President and Chief Information Officer); Boaz Efroni Rotman (Director of AI Strategy and Roadmap, Sales and Marketing Group (Oct 2025)). Ownership of AI on the org chart: CIO Cindy Stoddard runs Intel IT, reporting to CEO Lip-Bu Tan.
- What AI platform does Intel use?
- Intel IT built iForge, the back end that connects models and data sources, and iGPT, a no-code front end on which employees had built more than 7,000 personal assistants by Jan 2025. 1AI, an agent platform first deployed in Sales and Marketing, served more than 2,000 users in Oct 2025.
- Which AI models does Intel use?
- Intel runs open models such as Llama 3 70B, Qwen3 and Mistral 7B on its own Xeon and Gaudi hardware and sends complex, low-volume reasoning to frontier cloud models. Intel added Google's Gemini Enterprise in Jul 2026 and built its Ask Intel support assistant on Microsoft Copilot Studio.
- How many Intel employees use AI?
- More than 30,000 Intel employees used iGPT each month as of Jan 1, 2025, out of about 100,000 at the time. Intel had 85,100 employees at Dec 27, 2025 and has published no newer usage count.
- What return has Intel reported from AI?
- Intel has reported no business return that meets the ROI rule, for any use case or for the company as a whole. Intel IT's 2024-25 Annual Performance Report (Jun 2025) estimated $305 million of business value from AI use cases in 2024. The $305 million figure blends productivity, cost reduction, cost mitigation and incremental revenue with no published method, and Intel's figures for facilities, IT support and data centers are hours saved, workload, capacity or forecasts. Intel's rollout record puts it at Piloting. The same report counted more than 30,000 monthly users of iGPT, the company-wide generative AI platform Intel IT runs, as of Jan 1, 2025. CIO Cindy Stoddard has owned Intel IT and its internal AI-enabled transformation since Dec 1, 2025, reporting to CEO Lip-Bu Tan. That rollout evidence leaves the 18-month window in Dec 2026 unless Intel publishes a newer count. Intel falls short of Scaling because it has published no AI-credited return in cost saved, revenue, headcount or output per employee in any function, and no firm-level AI figure measured in those terms.
Open questions
- How does Intel IT measure the estimated $305 million of AI business value for 2024, and how much of it is hours saved converted into dollars? Why is the figure far below the more than $1.4 billion Intel IT credited to AI in 2023?
- Did Intel's AI use cases deliver the $376 million Intel IT projected for 2025, and will Intel publish a 2025-26 IT Annual Performance Report or another firm-level AI figure?
- How many Intel employees use iGPT, 1AI or Gemini Enterprise each week today, and will Intel publish a newer count, given that its newest figure, more than 30,000 monthly iGPT users as of Jan 2025 (reported Jun 2025), is the Piloting evidence and leaves the 18-month window in Dec 2026?
- What dollar value does Intel put on the capital purchases avoided by its design data center agents, and on the hours saved by its facilities, IT support and infrastructure agents?
- What do Ask Intel, the 1AI sales agents and the Accenture marketing arrangement return in cost or revenue, and how many marketing and support roles did they replace?
- How will Gemini Enterprise and Intel IT's own iGPT and 1AI platforms divide the work, and which frontier models does Intel use alongside its open models?
- Who owns internal AI on Intel's executive team beyond the CIO, given that Lip-Bu Tan's direct leadership of the AI groups after Sachin Katti's exit covers products, not internal tools?
- Does Intel use AI agents in its own chip design and verification flows, as its Google Cloud announcement and its EDA partners suggest, and with what effect on design cycle time or cost?
Sources
Intel
- Intel and Google Cloud collaboration on Intel's AI-enabled enterprise transformation (Jul 16, 2026)
- Intel appoints Cindy Stoddard as SVP and CIO (Nov 18, 2025)
- Intel announces leadership appointments (May 4, 2026)
- IT@Intel white papers index
- 2024-25 Intel IT Annual Performance Report (Jun 2025)
- 2023-24 Intel IT Annual Performance Report (Apr 2024)
- IT@Intel: Democratizing the use and development of generative AI across Intel (May 2024)
- IT@Intel: Validating and evolving Intel IT's multicloud strategy (Mar 2025)
- IT@Intel: Agentic AI in the enterprise (Oct 2025)
- IT@Intel: Agentic AI is transforming Intel facilities operations (Jan 2026)
- IT@Intel: Cost-effective and scalable infrastructure for GenAI workloads (Feb 2026)
- IT@Intel: Scaling a strong data foundation to an AI-ready data estate (Apr 2026)
- IT@Intel: Data center strategy leading Intel's business transformation (Jun 2026)
- IT@Intel: An AI-driven framework for enterprise data asset optimization (Jul 2026)
- IT@Intel: Transforming infrastructure management with agentic AI (Sep 2026)
- Intel Foundry and Formant: Proving the value of context in AI-driven industrial operations (Oct 2026)
Filings and reports
News
- CIO Dive: Intel taps Adobe exec as CIO amid turnaround efforts (Nov 19, 2025)
- The Register: Intel CTO and AI boss quits to join OpenAI (Nov 11, 2025)
- Tom's Hardware: Intel shifts customer support to an AI assistant built on Copilot Studio (Feb 22, 2026)
- Tom's Hardware: Intel to outsource marketing to Accenture and AI (Jun 21, 2025)
- Pure AI: Intel expands Google Cloud partnership as enterprise AI moves from pilots to core operations (Jul 17, 2026)
- Fast Company Middle East: Intel's chief people officer on AI (Mar 19, 2024)
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
Intel profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 9, 2026
