Why the old committee structure stopped working
IAPP's new Digital Risk Report, built from survey data, interviews with senior leaders, and analysis of securities filings, documents a shift that has been building for a while inside large organizations and is now showing up as formal, board-level strategy rather than an informal workaround. Organizations are moving away from separate governance committees for privacy, AI, and security toward what IAPP calls an aligned approach, a single framework with defined leadership, clear reporting lines, and explicit conflict-resolution mechanisms spanning all three domains at once, instead of three domains each defending their own narrow mandate.
Joe Jones, IAPP's director of research and insights, frames the underlying problem well: no regulation and no risk related to technology exists in a vacuum anymore, and treating them as though they did is what produced the fragmented governance structures most large enterprises are now trying to unwind. A privacy rule, a security control, and an AI model's data requirements now intersect constantly across nearly every new system an enterprise builds or buys. Treating them as three separate governance tracks, each with its own committee and its own sign-off process, means the organization typically only discovers the conflict between them when something breaks in production, rather than catching it earlier when a policy is first being written and the cost of fixing it is still low.
The committee sprawl enterprises actually built
The irony IAPP's report surfaces is that many organizations already tried to solve this by adding more committees, not fewer. Cross-domain committees multiplied as AI governance became a board-level concern, with privacy, security, legal, and AI ethics each staking a claim to oversight, often with overlapping membership and no clear sense of which group's decision actually wins when two of them disagree. Jones's description of the result, death by committee, is the kind of phrase that lands because every CIO reading it has sat through the overlapping meetings it describes, and has probably sat through several of them in the same week without anyone noticing the duplication.
The specific conflict IAPP highlights as a test case is instructive: data minimization requirements, a core privacy principle, run directly against an AI model's need for more data to perform well. In a siloed governance structure, the privacy committee and the AI committee can each approve their own side of that tension without ever resolving it, leaving engineering teams to quietly pick a winner in production, usually under deadline pressure and without documenting why. An aligned framework forces that resolution to happen as a deliberate decision, made by an accountable owner, before the conflict ever reaches production code rather than being discovered there by accident months later.
The talent gap is the real bottleneck
IAPP is direct about what is slowing this convergence down: a shortage of qualified personnel, and real difficulty attracting and retaining the specific people who combine privacy, security, and AI expertise in one role rather than specializing narrowly in just one of the three. The report warns this scarcity can leave genuine gaps in governance coverage precisely because innovation inside the business is outpacing the internal capacity available to oversee it responsibly, a mismatch that tends to widen further every quarter a company delays hiring for the combined role it actually needs rather than three separate specialists who each see only part of the risk.
That staffing reality explains why so many enterprises default to adding committees rather than consolidating them into one accountable structure. Appointing a cross-functional committee is organizationally cheap and can be done in a single meeting. Hiring or developing the small number of people who can actually own an integrated privacy-security-AI governance function is expensive and slow by comparison, and most enterprises have not budgeted for it at the scale the convergence IAPP describes actually requires over a multi-year horizon. The Schellman finding that only 27 percent of companies report a fully mature AI governance program is the direct, measurable consequence of that staffing gap, not a separate or unrelated data point.
Agents need governance that flexes, not governance that blankets
Gartner's research adds a practical constraint that strengthens IAPP's case for convergence. AI agents operating with different levels of autonomy across different tasks need governance that is adaptable and proportional to that autonomy, not a single blanket policy applied uniformly. An agent handling low-stakes scheduling needs a different permission structure than one authorized to move money or alter customer records, and Gartner identifies getting that distinction wrong as a leading cause of failed agent projects.
A siloed governance structure struggles to produce that kind of proportional policy because no single committee owns the full picture of what an agent actually does across its lifecycle, from the data it reads at the start of a task to the system it writes to at the end. An aligned framework, with one accountable structure spanning privacy, security, and AI, is far better positioned to calibrate permission levels against actual risk, because the people setting policy can see the whole operational picture rather than their domain's narrow slice of it, and can adjust an agent's autonomy up or down as its track record in production actually warrants.
What this means for the transformation roadmap
For CIOs building out AI governance this budget cycle, the practical move is to stop treating privacy, security, and AI oversight as three separate line items with three separate owners. Consolidating them into one framework with a single accountable leader is a governance decision, not a reorganization exercise, and it should be made before the next wave of agentic AI projects multiplies the number of permission decisions that need resolving.
The build-versus-buy question here is really a staffing question. Enterprises without the internal talent to run an integrated governance function should look seriously at external partners who can stand up that structure quickly, rather than defaulting to another internal committee that adds a meeting without resolving the underlying conflict. With AI agent deployments only accelerating, the cost of waiting to make this decision is measured in production incidents, not missed deadlines.



