Citation·page:schmidt-futures:fn55

Schmidt Futures - Footnote 55

Verdictcontradicted50%

1 check · 4/3/2026

WRONG NUMBERS: The source does not mention the grant amount of $500,000. FABRICATED DETAILS: The source does not mention Daniel Kang's work on AI agent cybersecurity attack benchmarks or risks like AI systems escaping laboratory control and self-replication. FABRICATED DETAILS: The source does not mention Bo Li's funding for developing virtual environments with red teaming algorithms to evaluate appropriate access levels for AI systems.

Our claim

entire record

No record data available.

Source evidence

1 src · 1 check

www.schmidtsciences.org/new-10-million-ai-safety-science-program-launched-for-foundational-research/

contradicted50%Haiku 4.5 · 4/3/2026

NoteWRONG NUMBERS: The source does not mention the grant amount of $500,000. FABRICATED DETAILS: The source does not mention Daniel Kang's work on AI agent cybersecurity attack benchmarks or risks like AI systems escaping laboratory control and self-replication. FABRICATED DETAILS: The source does not mention Bo Li's funding for developing virtual environments with red teaming algorithms to evaluate appropriate access levels for AI systems.

Case № page:schmidt-futures:fn55Filed 4/3/2026Confidence 50%