Chief AI Officer: Scope and Mandate

Last verified: September 20, 2026. A Chief AI Officer (CAIO) is the executive a company names to own its AI strategy, its AI deployment decisions, or both — but six SEC filings from April 2025 through September 2026 show no single template for what that means. Some companies gave it a stand-alone reporting line to the CEO. Others folded it into an existing technology title. Two used it only as a credential in a director’s outside job description. This guide walks through all six, in their own words.

This page reads corporate SEC filings for what they actually say about a job title. It is not legal, compensation, or investment advice, and it is not a claim about what any company is required to do.

The Six Filings at a Glance

Three companies appointed their own Chief AI Officer. One combined the title with an existing role. Two disclosed the title only as a sitting director’s job at a different company. All six are primary-source SEC filings, not secondary commentary.

Company	Filing	Date filed	Person	What it actually shows
FactSet Research Systems (NASDAQ: FDS)	8-K, Item 5.02	2026-03-04	Kate Stepp	Own-company CAIO, moved from CTO
Stagwell Inc. (NASDAQ: STGW)	8-K, Exhibit 99.1	2025-04-02	John Kahan	Own-company, inaugural CAIO
Resources Connection / RGP (NASDAQ: RGP)	8-K, Item 5.02	2026-09-03	Jessica Block	Own-company CAIO, moved to interim CFO
Unity Software (NYSE: U)	DEF 14A proxy	2026-03-27	Felix Thé	Combined title: SVP, Chief AI Officer and Product and Technology, Growth
Xerox Holdings (NYSE: XRX)	DEF 14A proxy, director bio	2025-04-09	John Roese	Outside credential — Roese is Dell’s CAIO, disclosed in a Xerox director’s bio
United Therapeutics (NASDAQ: UTHR)	DEF 14A proxy, director bio	2026-04-29	Ray Kurzweil	Outside credential — Kurzweil is Beyond Imagination’s CAIO, disclosed in a UTHR director-nominee bio

Where a Company Named Its Own Chief AI Officer

Three of the six filings are a company creating the role for itself, with an 8-K disclosing the appointment under Item 5.02 (officer changes).

FactSet Research Systems: a CTO moved sideways into the role

Company: FactSet Research Systems Inc. (NASDAQ: FDS).
Filing: Form 8-K, accession 0001628280-26-014385, filed March 4, 2026, reporting an event dated March 2, 2026 (Items 5.02 and 9.01).
Person: Kate Stepp, previously FactSet’s Chief Technology Officer since September 2022.
Reports to: Sanoke Viswanathan, Chief Executive Officer.
Verbatim scope, from the filing’s press-release exhibit: Stepp will “lead FactSet’s AI services for clients,” with a focus on “driving the development and deployment of AI capabilities across the company’s products and client solutions, and partner across the organization to further accelerate the roll out of AI solutions in each client type.”
What happened to the CTO role: it didn’t stay vacant. Bob Stolte became CTO the same day, also reporting to Viswanathan, with a scope the filing describes as “FactSet’s global technology organization, with responsibility for platform modernization, engineering execution, operational resilience, including system reliability, cybersecurity, IT business continuity, and enterprise technology strategy.” FactSet split what had been one executive’s job into two: an AI-facing role and an infrastructure-facing role, both reporting directly to the CEO.

Stagwell: an explicitly “inaugural” appointment

Company: Stagwell Inc. (NASDAQ: STGW), a marketing and communications holding company.
Filing: Form 8-K, accession 0001104659-25-030920, filed April 2, 2025, with the announcement in Exhibit 99.1.
Person: John Kahan, described in the filing as “the network’s inaugural Chief AI Officer” — Stagwell’s first person to hold the title.
Reports to: Mark Penn, Chairman and Chief Executive Officer.
Verbatim mandate, from the filing: Kahan “will spearhead the integration and development of artificial intelligence across Stagwell’s global network.”
Background disclosed in the filing: nearly four decades in data and artificial intelligence, with prior tenures at Microsoft and IBM.

Resources Connection (RGP): a CAIO appointment that became a launchpad

Company: Resources Connection, Inc., doing business as RGP (NASDAQ: RGP), a consulting and staffing firm.
Filing: Form 8-K, accession 0001084765-26-000059, filed September 3, 2026, reporting an event dated September 1, 2026.
Person: Jessica Block, who joined RGP as Chief AI Officer in March 2026.
What the same filing discloses: Block moves into the role of interim Chief Financial Officer effective October 2, 2026, when Jennifer Y. Ryu’s resignation as CFO takes effect. RGP’s own language: her experience “leading and scaling global businesses, combined with her technology expertise, positions her to support RGP through this transition and advance the company’s AI strategy.”
Background disclosed in the filing: before RGP, Block held a Factor (a legal managed-services firm) role leading “AI transformation and the global client delivery organization,” and had separately stepped in as Factor’s interim CFO too. She also held senior roles at Ankura Consulting and spent over a decade at FTI Consulting.
Why this one is worth flagging on its own: in this single filing, the CAIO title did not function as a terminal destination. It was one stop, about six months long, on the way to a CFO seat — and it was the second time in Block’s career that an AI-transformation role preceded an interim-CFO move at the same company. That is a pattern of one person’s career, not evidence about how the CAIO title behaves generally; the other five filings here show no comparable move.

When “Chief AI Officer” Is Folded Into an Existing Title

Not every company that discloses a “Chief AI Officer” created a stand-alone C-suite seat for it. Unity Software’s proxy shows the opposite: the title layered onto a job that already covered other ground.

Company: Unity Software Inc. (NYSE: U).
Filing: DEF 14A definitive proxy statement, accession 0001810806-26-000021, filed March 27, 2026.
Person and exact title, as disclosed: Felix Thé, “SVP, Chief AI Officer and Product and Technology, Growth.”
What that title structure shows: at Unity, “Chief AI Officer” is one clause inside a compound SVP title that also covers product and growth-technology responsibility — not a separate reporting line the way FactSet, Stagwell, and RGP structured it.

Xerox Holdings’ 2025 proxy shows a related but distinct pattern, covered next: the same combined-title structure, but at a different company than the one filing the disclosure.

When “Chief AI Officer” Is an Outside Credential, Not an Internal Hire

Two of the six filings don’t disclose a company’s own CAIO at all. They disclose a board director’s job title at a different company, inside that director’s biography. Both are worth reading precisely, because it is easy to misread either one as the filer having its own Chief AI Officer.

Xerox Holdings: a director whose day job is Dell’s CAIO

Company disclosing this: Xerox Holdings Corp (NYSE: XRX).
Filing: DEF 14A definitive proxy statement, accession 0001770450-25-000017, filed April 9, 2025.
Where it appears: the director biography for John Roese, a Xerox board member.
Verbatim description: Roese “has served as President and Global Chief Technology Officer of Dell Technologies since 2019” and is “also Dell’s Chief AI Officer responsible for driving Dell’s AI strategy and internal use of AI.”
What this is and isn’t: Roese’s CAIO title belongs to Dell Technologies, not to Xerox. Xerox’s proxy discloses it because it is relevant experience for a board nominee, the same way any director bio lists an outside executive role. Xerox itself is not shown appointing a Chief AI Officer in this filing.

United Therapeutics: a director-nominee whose CAIO title is at a company he co-founded

Company disclosing this: United Therapeutics Corporation (NASDAQ: UTHR).
Filing: DEF 14A definitive proxy statement, accession 0001082554-26-000009, filed April 29, 2026, for a June 26, 2026 annual meeting.
Where it appears: the Director Nominees section, in Ray Kurzweil’s listed title.
Verbatim description: Kurzweil is listed as “Chief AI Officer and Co-Founder, Beyond Imagination.”
What this is and isn’t: Beyond Imagination is a robotics company Kurzweil co-founded; his CAIO title belongs there, not to United Therapeutics. UTHR’s proxy lists it as his current professional role, the standard content of a director-nominee biography — it is not United Therapeutics disclosing an internal AI-officer appointment.

Put together, these two filings show something the three internal appointments don’t: “Chief AI Officer” has become a title specific enough, and recognizable enough, that proxy statements now use it as shorthand for a director’s outside expertise — the same way a bio might say “Chief Financial Officer” or “General Counsel” at another company.

Where NIKOLAI Fits In

CASRAI’s own NIKOLAI project — an independent, unendorsed reference dictionary of frontier-AI-safety terms, not a governance standard endorsed by any of the companies or filings above — has an element built for exactly the governance function these six filings are, in different ways, instantiating. Verified directly against NIKOLAI’s live element page on September 20, 2026, Accountable Decision-Maker and Sign-Off (element N9, Commitments and Governance track, status: proposed, nikolai-v0.1) defines the record as capturing “the named role (and, where published, the named person) who makes or approves a threshold determination, risk-acceptance decision, deployment decision, redaction, or framework change, together with the approval record itself (what was approved, by whom, and when).” NIKOLAI’s own page is explicit that this is “CASRAI’s own proposed definition, not a definition any named organisation has agreed to.”

The connection is direct rather than inferred: the same element page’s crosswalk section — NIKOLAI’s own reading of published frontier-AI-developer frameworks, not a mapping any of those developers has declared — records that Meta names “Chief AI Officer” as the primary accountable decision-maker role in its own framework (with “Director of Alignment and Risk” as an alternate under its §2.1.3), while other labs use different named roles: Anthropic points to its CEO and a designated Responsible Scaling Officer, OpenAI to its CEO or a person the CEO designates, Cohere to its Chief Scientist acting under delegated CEO authority, Microsoft to “Executive Officers responsible for AI governance,” and Google DeepMind and xAI to governance functions or “risk owners” described without a single named title. Meta’s own choice of “Chief AI Officer” as that named role, inside a frontier-model safety framework, is the same title this guide has just traced through six SEC filings for a broader kind of corporate AI governance — different regulatory context (frontier-model deployment decisions versus enterprise AI strategy and adoption), same underlying idea: name a specific role, and record what that role signs off on. None of the six SEC filers above are shown declaring alignment with NIKOLAI, and NIKOLAI’s element is not evidence that any of them modeled their CAIO role on it — the parallel is structural, observed independently from both sides, not a claim about cause or coordination.

Frequently Asked Questions

Is “Chief AI Officer” a legally defined or required title?

No. Neither federal law nor any state AI statute, including California’s SB 53 (the Transparency in Frontier Artificial Intelligence Act), requires a company to appoint a Chief AI Officer or defines what the title must mean. SB 53 requires a covered frontier AI developer to describe its internal governance practices in a published framework, but it does not name a required role or title — a distinct but related point covered in CASRAI’s guide to accountable decision-makers under SB 53. Every appointment in this guide is a voluntary corporate governance choice, not a statutory one.

Does a Chief AI Officer replace the CTO or the CISO?

It depends on the company. FactSet and Stagwell created the CAIO as a distinct seat reporting straight to the CEO, with the CTO role continuing separately (FactSet appointed a new CTO, Bob Stolte, the same day it named Kate Stepp CAIO). Unity Software instead folded the title into an existing role, “SVP, Chief AI Officer and Product and Technology, Growth.” Dell Technologies (disclosed via a Xerox director bio) combines it with the CTO title in a single person, John Roese. None of the six filings here show a CAIO absorbing or replacing a Chief Information Security Officer role.

Who does a Chief AI Officer typically report to?

In the two filings that disclose a reporting line for an internally appointed CAIO, it is the CEO directly: Kate Stepp reports to FactSet CEO Sanoke Viswanathan, and John Kahan reports to Stagwell Chairman and CEO Mark Penn. The RGP and Unity filings do not disclose an explicit reporting line for the role.

Is the Chief AI Officer role a permanent C-suite fixture once created?

Not necessarily, based on the one case here where it’s possible to check: RGP’s Jessica Block moved from Chief AI Officer into interim CFO about six months after joining in the CAIO role, per RGP’s own September 2026 filing. That is evidence about one filing, not a general finding about how the title behaves across companies — the other five filings in this guide show no comparable transition.

Does naming a Chief AI Officer satisfy an AI governance regulation like SB 53?

No, and none of the six filings claim it does. SB 53’s governance requirements attach to a covered developer’s frontier AI framework and its internal processes, not to any specific job title. A company could name a Chief AI Officer without that framework existing, or maintain a compliant framework without using that title at all; the two are governed separately.

Why This Matters for Research Administration

University IT-governance and research-computing offices face the identical structural question this guide traces across six SEC filings: does “who owns AI decisions” get a stand-alone named role, get folded into an existing CIO or VP-for-Research title, or stay informally distributed with no single accountable owner? This guide’s own three-way split is a usable menu for a provost or CIO deciding how to structure AI governance across a research enterprise. The parallel is sharpened by where it breaks down: a university’s version of this decision carries added layers — human-subjects oversight of AI-assisted research methods, and sponsored-research compliance — that none of the six SEC filers have to reconcile inside a single named role.