AI Absorption Ledger / FDX

FedEx

FDX · Q2 2026 · reported 2026-06-23 · revenue $25.01bn

Assessment

AI appears in the fiscal 2026 10-K far more than on either call. The 10-K says the digital transformation is underpinned by the digital backbone and AI capabilities with AI-enabled workflows that lower the cost to serve (claim c1), that the company is increasingly using AI in its operations (claim c3), and that it runs an AI policy, an AI Council and an enterprise-wide AI Literacy Program (claims c5 and c6); it describes FedEx Virtual Assistant, Tracking+ and Returns+, and a new AI assistant in the Developer Portal. The call adds the CEO’s remark that AI is embedded in the DRIVE process and the Chief Customer Officer’s AI and data center demand. The quarter opens the workflow, Virtual Assistant, developer assistant, data-centre demand and spending channels; none carries a dollar, bound or share from management.

The spending and Virtual Assistant channels are sized, both by the ledger: income-statement AI spending of , a share of revenue taken from the ledger’s own estimates for other large buyers, and of customer service cost the Virtual Assistant may have avoided against the prior-year rate, with a range that allows no saving at all at the low end. The AI-enabled workflows and the data-centre demand stay unsized because AI is named beside other causes and nothing separates its part; the products for merchants have no stated price; the developer assistant has no base. The anchor already described AI-enabled robotic sortation tests, the Virtual Assistant, AI-enhanced tracking and AI and machine learning behind DRIVE, so the robots, the workflows and the Virtual Assistant are read as relabelled and add nothing to the incremental total. Tracking+ and Returns+ are read as expanded: new AI products beside FedEx Returns Technology, which the 10-K still describes under its own name. The partnerships with ServiceNow and Dun & Bradstreet are context, not a channel: announced, with no launch, customer, price or term, and no stated direction of money (claim c13).

Every measured saving is credited elsewhere: transformation-related savings beat the target for fiscal 2026, and Network 2.0 and associated One FedEx savings are on track for by the end of calendar 2026, none of it attributed to AI. Salaries and employee benefits moved by on a year earlier, explained by wages, incentive compensation, benefits and exchange rates. The robots named in Q3 FY2026 are not mentioned, and the 10-K now describes robotic sortation tests without the artificial intelligence-enabled wording the anchor used.

FedEx Freight was spun off on 2026-06-01, after the quarter ended, so this quarter and Q3 FY2026 still consolidate it and compare with each other and with the prior-year quarters. The 8-K of 2026-07-21 is not an earnings release for this quarter: it recasts calendar 2024 and 2025 quarters with Freight as discontinued operations and new segments, it names no AI, and it is read for context only: no figure or claim is taken from it. From the transition period that starts on 2026-06-01, the reported lines exclude Freight and follow a calendar year, so the next quarter does not compare with this one without the recast.

Sized channels against the income statement, Q2 2026

2 of 7 channels sized

Each blue mark is one channel's dollars for the quarter; a bar is the range of an estimate. Grey marks are the company's reported lines. The distance between them is the point: how large the AI channel is next to the line it sits in.

New money and old money, Q2 2026

4 expanded3 relabelled

Each channel is tagged once for whether its money existed before language models, from the company's annual report and call at the start of the period. A bar splits one flow's sized dollars by that tag. The incremental total is the part that would not be there without the models: a new channel counts in full, an expanded one only for what AI added, a relabelled one at zero.

Paid for AI$2.5mn to $96mn sized

expanded $2.5mn to $96mn

Incremental total $0 to $96mnpoint $0$10mn in 1 channel has no traced baseline
Cost displaced by AI$0 to $17mn sized

expanded not sizedrelabelled $0 to $17mn

Incremental total $0

The sized total counts every channel the company credits to AI, including relabelled money that existed before language models and the ledger's own estimates for it. The incremental total counts relabelled channels at zero. A flow is split by layer where its dollars sit at more than one: end use, compute sold to builders, and hardware. The same dollar can be a buyer's spend, a cloud's revenue and a chipmaker's revenue, so the layers are never added together.

Paid for AI1 channel · $2.5mn to $96mn sized · $0 to $96mn incremental

engineering

Investment in AI capability and adoption

0.01% to 0.38% of the quarter’s revenue

Incremental total: counts at zero at the point and in full at the high end, with no traced baseline.

our inferencedisclosure: described· motive: exploratory· before LLMs: expanded

The 10-K says the company increasingly uses AI in its operations and AI-enabled technologies of its vendors, and names an AI policy, an AI Council, an AI Literacy Program for all team members and its investments in AI technologies; no amount is given and the call is silent on spending. The size shown is the ledger’s own estimate of income-statement spend, , a share of revenue taken from the ledger’s own estimates for other large buyers, which rest on judgment shares rather than disclosures. Expanded and sized as a level with no traced baseline, it is undetermined in the incremental total.

Evidence: 6 quotes, 3 figures, 3 confounds, 1 from before coverage

What the company spends building, buying and adopting AI. The 10-K says it increasingly uses AI within its operations and AI-enabled technologies of third-party vendors and service providers, names its investments in AI technologies, an AI policy and an AI Council, and an enterprise-wide AI Literacy Program launched in 2026 for all team members. No amount is given, nor how much is capitalized software; the 10-K explains other operating expense by outside service contracts and professional fees without naming AI. The counterparty is mixed: the company’s own technology staff and outside vendors and service providers, none named. Payments to the robotic pilot vendors, if expensed, sit here; no source separates them.

Why this motive

The 10-K names investments in AI technologies as a risk that may not pay off (claim c4) and describes an AI policy, an AI Council and an AI Literacy Program for all team members (claims c5 and c6), with no amount, return or line. Investing framing with no measure is the exploratory tell; training is not a usage mandate, so no narrative-defensive tell is quoted.

Before LLMs: expanded

At the anchor technology investment was already an activity: the 10-K credited data and technology, including artificial intelligence and machine learning, with supporting the DRIVE transformation (claim fdx-anchor-c1), and FedEx Dataworks had about employees. No AI share of any line was given then or since, so the level has no traced baseline. The size is the whole of an activity that existed before.

“Leveraging the capabilities of FedEx Dataworks, developments in data and technology, including artificial intelligence and machine learning, are facilitating the execution of our DRIVE transformation by creating new opportunities to improve our operational efficiency.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Capital expenditures, fiscal 2026 · FY2026
  • Capital expenditures on information technology, fiscal 2026 · FY2026
  • FedEx Dataworks employees, approximate · as-of 2026-05-31

What else could explain it

  • line composition: Other operating expenses rose on credit losses, outside service contracts, professional fees and exchange rates (claim c24), with no AI named; spin-off professional services and software licences sit in the same lines.
  • other: FedEx Dataworks had about employees at year end against at the anchor. No source explains the fall or ties it to AI, so the count is context and not a measure of AI spending.
  • bundling: AI is named inside a digital transformation that also covers the data platform and network systems; capital spending on information technology was in fiscal 2026 and total capital spending , shown as context and not as a base.

Quotes

“We are increasingly utilizing AI within our operations.”
c3 · Filing, risk factors, 10-K periodic report, 2026-07-20
“Further, our investments in AI technologies may not improve our services, operations, efficiency, or profitability to the extent anticipated and may divert resources from other strategic initiatives.”
c4 · Filing, risk factors, 10-K periodic report, 2026-07-20
“In 2026, we launched an enterprise-wide AI Literacy Program for all team members to provide knowledge and skills to use AI responsibly and effectively and to build AI confidence and capability across FedEx.”
c5 · Filing, business, 10-K periodic report, 2026-07-20
“We have an AI policy to support the responsible use of AI technologies in our operations, with a focus on enhancing business effectiveness while managing ethical, legal, cybersecurity, data privacy, and other technology-related risks. We also established an AI Council comprised of a cross-functional group of employees to support the responsible evaluation, governance, and use of AI technologies across the enterprise.”
c6 · Filing, risk factors, 10-K periodic report, 2026-07-20
“FedEx increasingly utilizes artificial intelligence-enabled technologies (“AI”) within its operations and also evaluates risks associated with the use of AI by third-party vendors and service providers.”
c7 · Filing, risk factors, 10-K periodic report, 2026-07-20
“Other operating expenses increased 8% in 2026 primarily due to increased credit losses, higher outside service contracts and professional fees, and unfavorable exchange rates.”
c24 · Filing, mdna, 10-K periodic report, 2026-07-20

Cost displaced by AI4 channels · $0 to $17mn sized · $0 incremental · 3 not sized

operations · cheap to verify

Robotic trailer loading and unloading (physical AI)

Not sized

The call and the release do not mention the robots, and the 10-K describes robotic sortation tests without naming AI (claim c22); the pilots named in Q3 FY2026 have no unit count, cost or saving, and their deployment beyond the pilots was dated to later in calendar 2026.

inscrutabledisclosure: not mentioned· motive: exploratory· before LLMs: relabelled

No source of the quarter names the Berkshire Grey or Dexterity robots. The 10-K still describes tests of driverless handling and robotic product sortation, in a sentence the anchor phrased with artificial intelligence-enabled sortation and the fiscal 2026 10-K phrases without it (claim c22), so the passage is not an AI claim. Salaries and employee benefits moved by on a year earlier, explained without AI.

Evidence: 0 quotes, 3 figures, 1 confound, 1 from before coverage

Package-handling labour that robotic trailer unloaders and loaders may displace, which the CEO calls physical AI on the fiscal Q3 2026 call: the Scoop robotic package unloader from Berkshire Grey and trailer-loading robots from Dexterity, both in the pilot phase, with further deployment expected later in calendar 2026. The cost would show in salaries and employee benefits and in purchased transportation. No source gives a unit count, a cost or a saving, and what the company pays either vendor is not stated. Network 2.0, DRIVE and the facility automation the 10-K describes are not attributed to AI and are confounds here.

Why this motive

Carried from Q3 FY2026, where the CEO described the robots as pilots (claim c2); the quarter is silent on them, and silence is not evidence about motive.

Before LLMs: relabelled

At the anchor Federal Express was already testing artificial intelligence-enabled robotic product sortation systems for small packages (claim fdx-anchor-c2), inside salaries and employee benefits of for fiscal 2024. The robots in coverage are still pilots and no line, price, volume or rate is shown to move because of them, so the activity is read as renamed.

“Federal Express is also testing autonomous, driverless technologies in the handling of large, non-conveyable packages, as well as artificial intelligence-enabled robotic product sortation systems to sort small packages.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Salaries and employee benefits · 2026-CQ2
  • Salaries and employee benefits, prior-year quarter · 2025-CQ2
  • Change in salaries and employee benefits on the prior-year quarter · 2026-CQ2

What else could explain it

  • transformation program: Fiscal 2026 transformation-related savings beat the target, and Network 2.0 and associated One FedEx savings are expected to reach by the end of calendar 2026; none of it is attributed to AI or to the robots.

By quarter

  • Q1 2026described · inscrutable · exploratory
  • Q2 2026not mentioned · inscrutable · exploratory

operations · cheap to verify

AI-enabled workflows in operations and the DRIVE process

Not sized

The 10-K credits the cost effect to the digital backbone and AI together (claim c1) and no source gives a saving, rate or line for AI’s part; every measured saving is credited to the transformation programs. A joint-cause channel nothing separates stays unsized, with total operating expenses quoted as the ceiling (the methodology rule for AI named beside another cause).

described, no sizedisclosure: described· motive: exploratory· before LLMs: relabelled

The 10-K says the digital transformation is underpinned by the digital backbone and AI capabilities, with AI-enabled workflows that drive faster response, fewer exceptions and a lower cost to serve, and that the company is increasingly using AI within its operations; the CEO says AI is embedded in the DRIVE process. No money is attached. The anchor already credited AI and machine learning with supporting DRIVE, so the channel is read as relabelled; it is left unsized, with total operating expenses of as the ceiling.

Evidence: 7 quotes, 8 figures, 3 confounds, 1 from before coverage

Operating cost avoided by AI-enabled workflows, which the 10-K credits, together with the company’s digital backbone, with faster response, fewer exceptions and a lower cost to serve, and which the CEO describes as AI embedded into the DRIVE process. No saving, rate or line is given for AI. Every measured saving is credited to DRIVE, Network 2.0, Tricolor and the international transformation programs, none of it to AI, so those programs are confounds here. Machine-learning volume forecasts and the FedEx Dataworks models the 10-K describes without naming AI are context, not this channel.

Why this motive

The 10-K credits AI-enabled workflows, together with the digital backbone, with faster response, fewer exceptions and a lower cost to serve, and the CEO says AI is embedded in the DRIVE process (claims c1 and c2), with no saving, rate or line; every measured saving is credited to the transformation programs (claims c19 and c20). AI named as a cause with no measure and no line moving is the exploratory tell; the efficiency tell is not met, since a result credited to several changes together does not meet it for any of them.

Before LLMs: relabelled

At the anchor the 10-K already credited developments in data and technology, including artificial intelligence and machine learning, with facilitating the DRIVE transformation and creating opportunities to improve operational efficiency (claim fdx-anchor-c1). No line, price, volume or rate is shown to move because of AI during coverage, so the activity is read as renamed.

“Leveraging the capabilities of FedEx Dataworks, developments in data and technology, including artificial intelligence and machine learning, are facilitating the execution of our DRIVE transformation by creating new opportunities to improve our operational efficiency.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Total operating expenses · 2026-CQ2
  • Total operating expenses, prior-year quarter · 2025-CQ2
  • Salaries and employee benefits · 2026-CQ2
  • Salaries and employee benefits, prior-year quarter · 2025-CQ2
  • Change in salaries and employee benefits on the prior-year quarter · 2026-CQ2
  • Transformation-related savings target for fiscal 2026, exceeded · FY2026
  • Network 2.0 and associated One FedEx savings expected by the end of calendar 2026 · as-of 2026-06-23
  • Network 2.0 and associated One FedEx savings expected by the end of calendar 2027 · as-of 2026-06-23

What else could explain it

  • transformation program: DRIVE, Network 2.0, Tricolor and the international transformation programs beat the fiscal 2026 savings target of (claim c19), and Network 2.0 and associated One FedEx savings are on track for by the end of calendar 2026 and by the end of calendar 2027 (claim c20); none of it is attributed to AI.
  • other: The 10-K credits the digital backbone and AI capabilities together (claim c1), and lists adopting AI and machine learning among the transformation initiatives (claim c23); nothing separates AI’s part from the data platform, standardization and network consolidation.
  • line composition: Salaries and employee benefits moved by on a year earlier, which the 10-K explains by wage rates, incentive compensation, benefits and exchange rates (claim c21); total operating expenses of are the ceiling on what the workflows could touch.

Quotes

“Our digital transformation is underpinned by our digital backbone and artificial intelligence (“AI”) capabilities, enabling intelligent orchestration, standardization, and digitized and AI-enabled workflows that drive faster response, fewer exceptions, and lower our cost to serve.”
c1 · Filing, business, 10-K periodic report, 2026-07-20
“We have embedded AI into our drive process, enhancing the rigor with which we work.”
c2 · CEO, prepared remarks, earnings call, 2026-06-23
“We are increasingly utilizing AI within our operations.”
c3 · Filing, risk factors, 10-K periodic report, 2026-07-20
“We may be unable to achieve the expected operational efficiencies and network flexibility, alignment of our cost base with demand, cost savings and reductions to our permanent cost structure, and other benefits from our transformation initiatives (including our technology infrastructure, adopting and utilizing AI and machine learning technologies).”
c23 · Filing, risk factors, 10-K periodic report, 2026-07-20
“This progress, combined with a sharp focus on structural cost reduction, enabled us to exceed the $1 billion transformation-related savings target that we shared at the start of the fiscal year.”
c19 · CEO, prepared remarks, earnings call, 2026-06-23
“We remain on track to achieve $1 billion of Network 2.0 and associated One FedEx savings by the end of this calendar year, and the full $2 billion by the end of CY 2027.”
c20 · CEO, prepared remarks, earnings call, 2026-06-23
“Salaries and employee benefits expense increased 8% in 2026 primarily driven by higher wage rates, variable incentive compensation, and employee benefit expenses, and unfavorable exchange rate impacts.”
c21 · Filing, mdna, 10-K periodic report, 2026-07-20

customer support · cheap to verify

FedEx Virtual Assistant for customer shipping questions

0% to 0.07% of the quarter’s revenue

Incremental total: counts at zero.

our inferencedisclosure: described· motive: exploratory· before LLMs: relabelled

The 10-K describes FedEx Virtual Assistant as an AI-enabled service that answers customer shipping questions; the description is unchanged from the anchor, so the channel is read as relabelled and adds nothing to the incremental total. The size shown is the ledger’s own: , the extra customer service cost the assistant may have avoided against the prior-year rate, a decomposition from revenue of in which every multiplier is ledger judgment. The range allows no saving at all at the low end, because nothing shows the assistant resolving more than a year earlier.

Evidence: 2 quotes, 3 figures, 2 confounds, 1 from before coverage

Customer service cost avoided because FedEx Virtual Assistant, which the 10-K calls an artificial-intelligence-enabled service on fedex.com, answers customer shipping questions and lets customer service representatives and sales professionals focus on higher-value interactions. The cost sits in salaries and employee benefits and in other operating expense; no contact volume, share handled, headcount or saving is given, and no source says whether any customer service work is outsourced.

Why this motive

The 10-K says the assistant answers customer shipping questions so representatives can focus on higher-value interactions (claim c8), the same words as at the anchor, with no handling time, resolution share, headcount or cost. An implemented tool with no measured result attributed to it does not meet the efficiency tell; AI named with no measure and no line moving is the exploratory tell.

Before LLMs: relabelled

At the anchor the 10-K described FedEx Virtual Assistant in the same words, as an artificial-intelligence-enabled service answering customer shipping questions (claim fdx-anchor-c3). The fiscal 2026 10-K repeats the description and shows no line, volume or rate moving, so the activity is read as renamed.

“FedEx Virtual Assistant on fedex.com is an artificial-intelligence-enabled service that provides answers to customer shipping questions, allowing our customer service representatives and sales professionals to focus on higher-value customer interactions.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Revenue · 2026-CQ2
  • Salaries and employee benefits · 2026-CQ2
  • Change in salaries and employee benefits on the prior-year quarter · 2026-CQ2

What else could explain it

  • line composition: Customer service staff sit inside salaries and employee benefits, which moved by on a year earlier for wage, incentive, benefit and exchange-rate reasons (claim c21); no customer service line is reported.
  • other: Digital self-service tools that predate the assistant (tracking, Delivery Manager, Picture Proof of Delivery) also take questions away from representatives, and the call credits signature proof of delivery with fewer customer service calls for shippers; none of that is the assistant.

Quotes

“FedEx Virtual Assistant on fedex.com is an artificial-intelligence-enabled service that provides answers to customer shipping questions, allowing our customer service representatives and sales professionals to focus on higher-value customer interactions.”
c8 · Filing, business, 10-K periodic report, 2026-07-20
“Salaries and employee benefits expense increased 8% in 2026 primarily driven by higher wage rates, variable incentive compensation, and employee benefit expenses, and unfavorable exchange rate impacts.”
c21 · Filing, mdna, 10-K periodic report, 2026-07-20

partner support · cheap to verify

AI assistant in the FedEx Developer Portal

Not sized

No source states what developer and integration support costs, who provides it or how many developers use the portal, and the assistant was introduced in the last month of the quarter. There is no base to apply a share to and no comparable reading in the ledger for an integration-support desk at a carrier, so no reference class exists.

inscrutabledisclosure: described· motive: exploratory· before LLMs: expanded

The 10-K says an AI-driven assistant in the FedEx Developer Portal, introduced in May 2026, helps developers find documentation, get technical guidance and integrate faster with FedEx digital services. Read as expanded (an LLM-era tool deployed into existing integration support) and sized as an increment, it is left inscrutable: nothing gives the size of the work it touches.

Evidence: 1 quote, 1 confound, 1 from before coverage

Integration-support work that the AI-driven assistant in the FedEx Developer Portal, introduced in May 2026, may take over: helping developers find documentation, get real-time technical guidance and finish integrations with FedEx digital services. No cost, usage or saving is given, and no source states what developer support costs or who does it.

Why this motive

The 10-K says the AI-driven assistant was introduced in May 2026 to help developers access documentation and accelerate integrations (claim c12), with no usage, measure or cost. A new tool with no measure is the exploratory tell.

Before LLMs: expanded

At the anchor the company already deployed software systems and internet-based applications so that customers could connect their own package data with FedEx delivery information (claim fdx-anchor-c7); the support behind those integrations sat in salaries and employee benefits of for fiscal 2024 and other operating expense, with no developer-support amount. An AI assistant deployed in that work during coverage is read as an expanded activity. The size is the change AI made, not the whole line.

“Additionally, software systems and internet-based applications are deployed to offer customers new ways to connect internal package data with external delivery information.”
Filing, business, 10-K periodic report, 2024-07-15

What else could explain it

  • other: Integration support is also carried by existing developer documentation, account teams and partners; any saving would sit inside salaries and employee benefits or other operating expense, neither of which isolates it.

Quotes

“In May 2026, we introduced an AI-driven assistant within the FedEx Developer Portal to help developers more efficiently access documentation, receive real-time technical guidance, and accelerate integrations with FedEx digital services.”
c12 · Filing, business, 10-K periodic report, 2026-07-20

Revenue arriving through AI2 channels · 2 not sized

product revenue · cheap to verify

Tracking+ and Returns+ post-purchase products for merchants

Not sized

No price, revenue, customer count or volume is given and no source says the products are priced apart from shipping, so they are read as launched products with no stated price (the methodology rule for products with no price): described and unsized, with no ballpark built on an invented price.

described, no sizedisclosure: described· motive: exploratory· before LLMs: expanded

The 10-K describes Tracking+ and Returns+ as AI-powered, white-labelled post-purchase solutions embedded in merchants’ channels that automate responses to common customer inquiries, and says adoption is scaling in the United States. No money is attached and the call is silent on them. Expanded, sized as an increment: the products add automated customer responses beside FedEx Returns Technology, the returns-management product the anchor shows and the 10-K still describes under its own name. With no stated price the channel stays unsized and adds nothing to the incremental total while it does.

Evidence: 2 quotes, 3 figures, 2 confounds, 2 from before coverage

Revenue from FedEx Tracking+ and FedEx Returns+ (Returns Plus on the call), which the company calls AI-powered, white-labelled post-purchase solutions for enterprise customers, embedded in merchants’ own digital channels: delivery and returns visibility, automated responses to common customer inquiries, proactive communications and analytics. Announced in January 2026 for the U.S., with Europe planned from April 2026. No price, revenue, customer count or volume is given, and no source says whether the products are priced apart from shipping. The verification is cheap because the product automates answers to routine delivery questions.

Why this motive

The 10-K describes the products as AI-powered and says they reduce merchants’ support demand and that adoption is scaling (claims c10 and c11), with no price, customer count, attach rate or revenue; the call does not mention them. No offensive tell is met, so the motive stays exploratory.

Before LLMs: expanded

At the anchor the company already sold FedEx Returns Technology, giving high-volume merchants visibility into returns and shipment tracking (claim fdx-anchor-c4), and credited an AI model with more accurate estimated delivery dates on fedex.com (claim fdx-anchor-c5). Tracking+ and Returns+ add automated responses to common customer inquiries beside that product, which the fiscal 2026 10-K still describes under its own name, so the activity is read as expanded and sized as an increment; no price, volume or revenue is given. The size is the change AI made, not the whole line.

“Additionally, FedEx Returns Technology, a comprehensive solution for returns management, provides high-volume merchants and e-tailers complete visibility into returns and an easy way to track shipments, manage inventory, analyze returns trends, and make more informed decisions based on shoppers’ returns behaviors.”
Filing, business, 10-K periodic report, 2024-07-15
“Additionally, an advanced machine learning and artificial intelligence model developed by FedEx Dataworks has enhanced tracking capabilities on fedex.com, delivering greater estimated delivery date accuracy, including updates for early or delayed shipments.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Revenue · 2026-CQ2
  • Revenue, prior-year quarter · 2025-CQ2
  • Consolidated revenue growth on the prior-year quarter · 2026-CQ2

What else could explain it

  • bundling: No source says whether Tracking+ and Returns+ are priced apart from shipping; their money may sit inside package revenue as part of the shipping relationship.
  • relabel: FedEx Returns Technology and AI-enhanced tracking estimates were already sold or described at the anchor (claims fdx-anchor-c4 and fdx-anchor-c5) and are still described in the 10-K beside the new products.

Quotes

“In January 2026, we announced FedEx Tracking+ and FedEx Returns+, AI-powered, white-labeled post-purchase solutions designed for enterprise customers and embedded directly within merchants’ digital channels.”
c10 · Filing, business, 10-K periodic report, 2026-07-20
“By integrating these capabilities into the post-purchase experience, these solutions help customers strengthen brand engagement, reduce support demand, and enhance overall customer satisfaction. We are scaling adoption in the United States and evaluating expansion opportunities internationally and across additional customer segments.”
c11 · Filing, business, 10-K periodic report, 2026-07-20

By quarter

  • Q1 2026described · described, no size · exploratory
  • Q2 2026described · described, no size · exploratory

customer cohort

Shipping demand from the AI and data-centre build-out

Not sized

AI is named together with data centres and industrial and power infrastructure throughout, the growth rate has no stated base or level, and the Chief Customer Officer declines to size the space (claim c17). Nothing separates AI’s part, so the channel stays unsized with consolidated revenue of as the ceiling (the methodology rule for AI named beside another cause and for data-centre demand).

described, no sizedisclosure: direction only· motive: exploratory· before LLMs: expanded· layer: facilities

The Chief Customer Officer calls the AI and data center space an emerging, rapidly scaling growth engine delivering double-digit revenue growth, from hyperscalers to the industrial and power infrastructure behind the build-outs, with wins especially out of Asia. Asked for a size, she says the company has started to size the market and gives no number. The state is directional (a rate, no level); the channel sits in the facilities layer, read as expanded and sized as a level with no traced baseline, and is left unsized. The counterparty is mixed: hyperscalers and their industrial and power-equipment suppliers, none named.

Evidence: 7 quotes, 5 figures, 3 confounds, 1 from before coverage

Package and air freight revenue from what the Chief Customer Officer calls the AI and data center space, which she says delivered double-digit revenue growth and spans hyperscalers and the industrial and power-infrastructure suppliers building data centres: time-critical moves of technology infrastructure, much of it out of Asia. No amount, share or base is given, and AI is named together with data centres throughout, so data-centre demand is a ceiling on the AI part and nothing separates AI’s part. The layer is facilities: the shipments are part of building and equipping data centres, where construction materials and power equipment sold into data-centre builds also sit; the hardware carried is another company’s revenue. The counterparty is mixed: hyperscalers and their industrial and power-equipment suppliers, none named.

Why this motive

The Chief Customer Officer gives a growth rate for the AI and data center space and says its growth rates are the highest of the key verticals (claims c14 and c16), but names AI together with data centres and industrial and power build-outs throughout and declines to size the space (claim c17). Measured movement credited to AI beside other causes does not meet the offensive tell for AI; with the tells in conflict the less durable motive is taken.

Before LLMs: expanded

At the anchor Federal Express already sold express freight services for the time-definite freight market (claim fdx-anchor-c6), with FedEx Express freight revenue of in fiscal 2024; no data-centre or AI customer was named. During coverage management gives a growth rate for the AI and data-centre space, so volume is read as having moved, and the channel is expanded and sized as a level with no traced baseline. The size is the whole of an activity that existed before.

“Federal Express also offers U.S. express overnight and deferred freight services to handle the needs of the time-definite freight market.”
Filing, business, 10-K periodic report, 2024-07-15

Figures

  • Revenue · 2026-CQ2
  • Revenue, prior-year quarter · 2025-CQ2
  • Consolidated revenue growth on the prior-year quarter · 2026-CQ2
  • Federal Express revenue growth on the prior-year quarter · 2026-CQ2
  • Fuel surcharge share of Federal Express revenue growth, percentage points · 2026-CQ2

What else could explain it

  • other: Data-centre demand is not an AI label: the Chief Customer Officer names hyperscalers beside industrial and power-infrastructure suppliers and an automotive customer that moved a line to power generation for data centres, and data centres built for general cloud and enterprise use need the same shipments; any data-centre amount would be a ceiling on the AI part. None is given.
  • mix shift: The Chief Customer Officer credits B2B growth first to profitable market share and adds some inventory buildup and restocking (claim c25); share gains across the key verticals explain part of the same growth.
  • other: Fuel: percentage points of Federal Express revenue growth of came from fuel surcharges (claim c26), so revenue growth overstates volume.

Quotes

“The AI and data center space is an emerging and rapidly scaling growth engine for us, delivering double-digit revenue growth. Rather than a narrow vertical, this space represents a horizontal ecosystem. We are capturing demand across the entire value chain, from traditional hyperscalers to the industrial and power infrastructure that support these massive build-outs.”
c14 · Executive, prepared remarks, earnings call, 2026-06-23
“Yes, we did like to see some of the wins from an AI and a data center perspective, especially coming out of Asia.”
c15 · Executive, qa, earnings call, 2026-06-23
“It is a little bit hard to categorize and define. What I can tell you is across both our industrial base and the AI base, it is growing. The growth rates are the highest in all four of the verticals.”
c16 · Executive, qa, earnings call, 2026-06-23
“Yes, we have started to size the total market, I will tell you what we have found is, that's why I mentioned in my prepared remark, is more of an ecosystem.”
c17 · Executive, qa, earnings call, 2026-06-23
“They have built incredible momentum, as I mentioned, in the data center and AI space, but also leaning into their industrial base out of APAC.”
c18 · Executive, qa, earnings call, 2026-06-23
“I do think that there's a little bit of inventory buildup and restocking going on.”
c25 · Executive, qa, earnings call, 2026-06-23
“This growth includes a 5 percentage point benefit from fuel price-driven surcharge revenue.”
c26 · Executive, prepared remarks, earnings call, 2026-06-23

Reported lines, year-over-year growth

Revenue +12.5%

Q2 2026. Growing slower than revenue: salaries and employee benefits (+11.4%). A displaced cost shows up as a line that stays under the dashed revenue line. These are the audited lines, as first reported; nothing here is attributed to AI by the filing.

Salaries and employee benefitsTotal operating expensesRevenue
-5%0%5%10%15%Q1 2025Q2 2025Q3 2025Q4 2025Q1 2026Q2 2026Total operating expensesRevenueSalaries and employee benefits
Reported values and filings
LineQ1 2025Q2 2025Q3 2025Q4 2025Q1 2026Q2 2026
Revenue$22.16bn$22.22bn$22.24bn$23.47bn$24.00bn$25.01bn
Salaries and employee benefits$7.88bn$7.69bn$8.06bn$8.39bn$8.82bn$8.57bn
Total operating expenses$20.87bn$20.43bn$21.06bn$22.09bn$22.65bn$23.46bn