Search

FTC investigation: Anthropic and OpenAI must testify about consumer damages

Translated title. Original: FTC-Untersuchung: Anthropic und OpenAI müssen zu Verbraucherschäden aussagen

The FTC has opened a consumer-damage investigation into AI developers, focusing on Anthropic and OpenAI, with subpoenas demanding extensive records and executive testimony. The move signals tighter regulatory scrutiny, potential costs for model design, and broader industry expansion of investigations, with guidance available on complying with AI regulations.

AI governance

AI governance defines policies, ownership, and controls for building and deploying AI, aligning internal practices with external frameworks like the EU AI Act, NIST RMF, and ISO 42001; it emphasizes inventory, risk classification, accountability, documentation, and ongoing monitoring.

AI governance requires integration across departments and new security controls in companies | TI INSIDE Online

Translated title. Original: A governança de IA exige integração entre áreas e novos controles de segurança nas empresas | TI INSIDE Online

AI governance must be embedded early and across departments, balancing privacy, security, and business needs. A multi-layered governance model, with business involvement and risk-based controls, is needed to scale responsible AI while managing Shadow IT and regulatory requirements.

How He Became the World’s Most Powerful Figure in AI

Translated title. Original: Slik ble han verdens mektigste innen KI

The article reports allegations of a relationship between Nvidia CEO Jensen Huang and a vice president who reports to him, which would breach company rules. It also says the FBI is investigating and that Donald Trump, described as Huang’s friend, had included him on a visit to China.

AI governance requires ongoing assessment and human oversight to reduce risks in companies | TI INSIDE Online

Translated title. Original: A governança de IA exige avaliação contínua e supervisão humana para reduzir riscos nas empresas | TI INSIDE Online

AI governance now requires ongoing, life-cycle supervision with guardrails and clear responsibilities; cross-functional use and startup collaboration raise security and compliance considerations, while human oversight remains essential for high-stakes decisions.

The Nvidia board’s artificial independence

Bored of governance

Multinational Governance Architecture | COMPEL Framework

The document presents a layered architecture for AI governance across jurisdictions, from shared global standards to local rules and technical controls. It explains how regulatory differences, cross-border data restrictions, and conflicting disclosure requirements complicate multinational compliance.

You Can’t Govern AI Without Governing The Information Beneath It

Governance of information underpins enterprise AI. The document argues that AI effectiveness depends on clear ownership, lifecycle management, and traceability of the data the AI uses, warning that information-governance debt can lead to wrong answers and compliance risks, with references to NIST and EU AI Act.

When AI gets it wrong: why governance can no longer be an afterthought

AI governance is moving from a technical concern to a boardroom priority, as regulators seek transparency and accountability; organisations must implement clear AI use policies, human oversight, and proactive risk management to mitigate hallucinations and regulatory risk while preparing for future regulation.

AI governance in the consortium market: eight pillars for mature adoption | TI INSIDE Online

Translated title. Original: Governança de IA no mercado de consórcios: oito pilares para uma adoção madura | TI INSIDE Online

AI governance in mature adoption: define eight pillars to align AI with business strategy, ensure accountability, manage risks, and secure data, with a focus on responsible, long-term value in sectors like consortia.