AI NOT SAFE YET, UCLA PROFESSOR WARNS

AI DESK■ 1 MIN READ

WED, JUN 10, 2026

■ AI-SUMMARIZED FROM 1 SOURCE ▸ TIMELINE

As AI companies Anthropic and OpenAI pursue IPOs on promises of safety, UCLA professor Safiya Noble counters that current AI systems remain unsafe due to embedded stereotypes and biases in training data.

Safiya Noble, director of UCLA's Center on Resilience & Digital Justice, challenged industry claims about AI safety during a Bloomberg Tech appearance. While major AI developers emphasize safety measures ahead of potential public offerings, Noble highlighted a critical gap: training data containing societal prejudices gets encoded into AI systems. The concern centers on how biases present in datasets used to train AI models propagate into deployed systems, potentially amplifying discrimination at scale. Noble's warning suggests that current safeguards fall short of addressing systemic bias issues baked into AI development. The remarks come as Anthropic and OpenAI both signal readiness for IPOs, with safety and beneficial AI use positioned as core value propositions to investors. Noble's critique indicates that technical safety protocols alone may not address deeper concerns about fairness and equity in AI systems.

■ SOURCES

► Bloomberg Tech

■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE

■ MORE FROM THE AI DESK

P848AI ADOPTION REMAINS SELECTIVE, NOT UNIVERSAL

Despite widespread AI hype, most people aren't integrating artificial intelligence into their daily routines. A recent analysis challenges the assumption that everyone is adopting AI tools across all aspects of their lives.

JUST NOW— AI Desk

P845RIO'S 'HOMEGROWN' LLM ALLEGEDLY A MERGED MODEL

Rio de Janeiro's claimed locally-developed large language model appears to be constructed from existing models rather than built from scratch, according to findings shared on GitHub.

JUST NOW— AI Desk

P836GOOGLE CLOUD LAUNCHES FORMAT FOR AI-READY KNOWLEDGE

Google Cloud introduced the Open Knowledge Format (OKF), a standardized specification that converts scattered organizational documents into portable Markdown files with YAML frontmatter. The format enables AI agents to access and process knowledge efficiently.

4H AGO— AI Desk

P835AI SLASHES LONDON FINANCE JOBS BY 77%

Finance analyst vacancies in London have plummeted from over 350 to around 80 in four years, according to recruitment data. The dramatic decline reflects AI's growing impact on professional roles in banking and corporate sectors.

6H AGO— AI Desk

◄ BACK TO NEWS

AI NOT SAFE YET, UCLA PROFESSOR WARNS

■ MORE FROM THE AI DESK

■ SUBSCRIBE TO THE DAILY BRIEF