Trustworthy AI Framework

Strategic Center Evaluation

Not part of the /25.

Adjacent Publisher Context

The following materials are published by Deloitte and address topics relevant to the Five Forces, but are not explicitly positioned as components of the Trustworthy AI™ Framework. They are not included in the /25 coverage score above.

Narrative Strengths & Gaps

Force 1 — Internal Pressures & Capabilities (4/5). This is the framework’s centre of gravity. It builds governance across the whole AI lifecycle and embeds it across technology, processes and employee training, as set out on the canonical page. Applying COSO ERM framework principles to AI connects the seven dimensions to an established enterprise risk management structure, including control activities such as inventorying, benchmarking and trend analysis. It also introduces the Omnia Trustworthy AI™ Module. Building Trustworthy Generative AI adds diverse governance teams and “regular waypoints for stakeholder review” across the GenAI lifecycle. The public materials work at principle and consideration level; the detailed, implementable tooling sits inside Deloitte’s engagement offerings, so Actionability is scored as directional. The framework does not assess where an organisation stands today (Map: Not met). Its lifecycle stages, from ideation to MLOps, give each AI system a governed path rather than a route for the organisation (Course: Partly met). Its destination is explicit: AI adopted with trust at the centre, safeguarded across seven dimensions.

Force 2 — Global & Geopolitical Dynamics (1/5). The framework recognises that “regulatory bodies worldwide” are setting GenAI rules. It also illustrates how synthetic media could misrepresent world leaders on the global stage (Building Trustworthy Generative AI). Sovereignty, export controls and cross-border data strategy are left to Deloitte’s broader research. That scope boundary is consistent with a framework built around trust in individual AI systems.

Force 3 — Societal & Regulatory Pressures (4/5). Ethics is the framework’s reason for being. The seven dimensions translate public and institutional expectations into design requirements, and the canonical page builds regulatory compliance into governance at every lifecycle stage. Human values in the loop supplies design principles. Toward humanity’s brightest future and the State of Ethics and Trust in Technology report extend this to societal impact and organisational ethics standards. The public guidance stays at the level of principles and considerations, so readers build the compliance programme themselves.

Force 4 — Industry & Competitive Pressures (1/5). The framework names pressure from competitors and start-ups as a driver of AI investment (Applying COSO ERM). It also states that accountability holds whether a model is built in-house or bought from a vendor (Building Trustworthy Generative AI). Market positioning and ecosystem strategy fall outside its trust-and-governance remit.

Force 5 — Human Needs & Aspirations (3/5). The framework speaks directly to external stakeholder expectations. Customers should be able to opt in and out of data sharing, and their data should not be used beyond its stated purpose (Applying COSO ERM). GenAI content must not harm customers or society (Building Trustworthy Generative AI). Public concerns about the future of work, equity and trust are framed as considerations that business leaders should weigh (Toward humanity’s brightest future). Talent-market expectations from prospective employees are not developed.

What this framework does not give you. A map of where the organisation stands. Without a clear picture of where it stands, an enterprise cannot tell which moves matter most. A route for the organisation as a whole; the lifecycle stages govern individual AI systems. Without a route, even a well-chosen destination stays an aspiration. A way to choose your own destination; it supports AI adopted with trust at the centre across seven dimensions. An enterprise that has not stated its own ambition cannot check whether this destination is the right one for it.

Synthesis. The Trustworthy AI™ framework serves Chief Risk Officers, Chief AI and ethics officers, internal audit, and boards most directly, especially in regulated organisations scaling AI that need a shared vocabulary for trust and a bridge into established ERM practice. Its destination fits organisations whose ambition includes trusted, ethically safeguarded AI. On the Strategic Center it stays implicit throughout: lifecycle stages rather than planning horizons, human-AI collaboration without an explicit choice model, and innovation balanced against ethics. The Five Forces lens adds a map of where the organisation stands, an organisation-wide route, a way to set the destination itself, and the open dimensions — geopolitical and competitive pressure, and the strategic questions of horizon, orientation and idea value — that sit beside a strong trust-and-governance core.

Sources

Tier 1 (scored), recency-gate pass unless noted:

1. Trustworthy AI™ framework — Bridging the ethics gap surrounding AI – 25 Jun 2025; canonical anchor
2. Building Trustworthy Generative AI – page updated 3 Jun 2025; underlying PDF dated Aug 2023, passes via active canonical link
3. Applying COSO ERM framework principles to AI – page updated 11 Jul 2025; underlying PDF dated Sep 2021, passes via active canonical link
4. Human values in the loop: Design principles for ethical AI – published 28 Jan 2020, page modified Jun 2026; passes via active canonical link
5. Toward humanity’s brightest future with Generative AI – page updated 9 Jul 2025; PDF Jun 2024
6. Preparing the workforce for ethical, responsible and trustworthy AI – 25 Mar 2025, updated 4 Jul 2025
7. State of Ethics and Trust in Technology (Deloitte’s 2024 ethical technology report) – 31 Mar 2025, updated 9 Jul 2025

Tier 2 (reported in Section 2.75, not scored):

1. Deloitte AI Institute hub – updated 25 Sep 2026; linked from the canonical page as navigation

Tier 3 (third-party corroboration, not scored):

1. None used.

Excluded on recency or quality (not scored):

1. Revised Draft of Colorado AI Regulation: GRMF summary (PDF) – Jul 2023; summarises a May 2023 draft that the final regulation has superseded
2. Five further research-agent links (three Deloitte service pages, a 2020 WSJ CIO Journal article and a 2020 press release) could not be opened on 24 Oct 2026 and are not cited.