AI Governance
Data Centers
Drones
Climate
Privacy
2026 Maine Senate
Bob Dylan
New York’s Synthetic Performers Law Draws First Complaints
Yesterday’s reporting offered two modest but concrete examples of AI governance becoming an operational matter. New York’s attorney general is reviewing four complaints alleging undisclosed AI-generated performers in advertising, while the mortgage industry has introduced a certification process intended to test whether AI suppliers can document their controls.
Neither development amounts to a wider regulatory realignment. But together they reinforce a pattern visible in recent briefings: the practical question is increasingly whether organizations can show how their AI systems are disclosed, governed, monitored and reviewed—not merely whether they have adopted high-level principles.
New York’s Synthetic Performers law has reached its first apparent compliance test. Mashable reported that the state attorney general received four complaints alleging that advertisers used undisclosed AI-generated models or performers in ads directed at New York consumers. The complaints are under review, not enforcement findings: the available reporting does not establish that any company violated the law, nor how the state will interpret its disclosure requirements. Still, complaint-driven scrutiny turns a recently enacted disclosure obligation into an immediate issue for advertisers using photorealistic synthetic people across state markets.
MISMO launched two AI-governance certifications for the mortgage sector: one for specific AI use cases in mortgage technology and another for advisory firms implementing governance programs. National Mortgage Professional reported that the assessments cover controls, documentation, testing, oversight, monitoring, evidence collection and executive reporting. The certifications are not regulatory requirements and their uptake is unknown. Their importance lies in making vendor assurance more concrete in a consequential lending market, where lenders need a workable basis to question suppliers about how a model is controlled rather than simply accepting policy statements.
Key Points
- Disclosure and assurance are becoming more tangible forms of AI oversight. New York’s process asks whether a visible AI use was properly identified to consumers; MISMO’s program asks whether less visible vendor practices can be documented and examined. They are different tools, but both shift attention from broad commitments toward evidence that can be reviewed after a complaint, during procurement or in a compliance assessment.
- Sector and state-level mechanisms continue to fill the practical governance space in the United States. Yesterday’s developments do not show a uniform national approach. Instead, they illustrate how obligations and expectations can arise through a narrow consumer-protection rule in one jurisdiction and a voluntary industry program in a regulated sector, leaving companies to manage distinct requirements by use case and market.
Implications
Advertisers using synthetic people should be able to identify where such material appears, determine when a state disclosure rule applies and preserve a clear response process if challenged. The New York complaints are unresolved, but they make it harder to treat disclosure as a purely theoretical requirement.
Mortgage lenders and AI suppliers should watch whether MISMO certification becomes a useful procurement or diligence signal. If lenders, Freddie Mac, Fannie Mae or supervisors begin to reference it, documented testing, monitoring and governance evidence could become more influential in vendor selection. Certification alone, however, does not establish regulatory compliance or eliminate liability.
Watchpoints
Watch
Whether New York’s attorney general opens a formal investigation, identifies respondents, issues interpretive guidance or seeks penalties under the Synthetic Performers law.
Watch
Whether mortgage lenders, Freddie Mac, Fannie Mae or relevant regulators incorporate MISMO’s certifications into vendor-review or oversight expectations.
Watch
Whether federal authorities move beyond existing case-by-case tools toward binding U.S. rules addressing AI-enabled political influence operations.
Fallout
The day’s clearest developments were early state-level scrutiny of synthetic advertising and a new sector-specific assurance mechanism for mortgage AI. Both point toward more practical compliance demands, but neither creates a broad new national obligation.
Synthetic-Performer Advertising Disclosure
New York’s Synthetic Performers law requires disclosures for covered uses of AI-generated performers in advertising directed at state consumers. Its early application matters beyond advertising because it offers a test of how state AI transparency rules may be enforced in practice.
Fresh developments
Mashable reported that New York’s attorney general has received four complaints alleging undisclosed AI-generated models or performers in advertising. The complaints remain under review, with no reported finding, penalty or authoritative interpretation of the law.
Why we noticed
The complaints are an initial indication that a disclosure law is moving from enactment into complaint-based oversight. Companies using photorealistic synthetic people now face a more immediate need to assess campaign disclosures and retain support for their compliance decisions.
Watch for:
- A formal investigation, named respondent, settlement, penalty or attorney general guidance clarifying what disclosure is required.
- Whether complaints expand beyond the currently reported allegations or prompt changes in advertising-platform policies.
Article links:
Mortgage AI Vendor Assurance
Mortgage lending combines high-consequence decisions, extensive third-party technology use and established expectations for documentation and oversight. That makes vendor governance a practical point of entry for AI assurance even where no new binding AI rule has been adopted.
Fresh developments
MISMO launched certifications for individual mortgage-technology AI use cases and for advisory firms that implement governance programs. The stated assessments cover controls, documentation, testing, oversight, monitoring, evidence collection and executive reporting.
Why we noticed
The initiative gives lenders and suppliers a sector-specific way to demonstrate governance practices. Its eventual importance will depend on adoption, but it reflects a broader move toward auditable evidence in vendor review rather than general AI-policy commitments.
Watch for:
- Whether lenders or major mortgage-technology suppliers begin using the certifications in procurement, diligence or marketing.
- Whether Freddie Mac, Fannie Mae or regulators treat certification as relevant evidence of vendor-governance practice.
Article links:
Final Thought
Yesterday’s movement was incremental, but its direction is clear: AI governance is increasingly being expressed through proof. A consumer complaint can test whether a disclosure was made; a lender can ask a vendor to show its controls. The next question is whether these mechanisms gain enough institutional backing to become routine expectations rather than isolated examples.