Skip to content
MONTEGRE
Sign in
BREAKING
/Mynt prices GCash IPO at P6.60, aims to raise up to P60.9B/Sergio Moro Wins Parana Governorship in First Round/Cleitinho Azevedo wins Minas Gerais governorship in first round/Pacific leaders open pre-COP summit in Fiji ahead of Antalya climate talks/Valkyries Beat Aces 71-60 in Semifinals Opener Behind Williams' 25 Points/López Aliaga leads Lima mayoral exit poll with 31.2%/Mynt prices GCash IPO at P6.60, aims to raise up to P60.9B/Sergio Moro Wins Parana Governorship in First Round/Cleitinho Azevedo wins Minas Gerais governorship in first round/Pacific leaders open pre-COP summit in Fiji ahead of Antalya climate talks/Valkyries Beat Aces 71-60 in Semifinals Opener Behind Williams' 25 Points/López Aliaga leads Lima mayoral exit poll with 31.2%

Ex-OpenAI Safety Lead Warns AI Firms Not Careful Enough

David Robinson says companies prioritize speed over safety and calls for external incentives to slow development.

Sources: Business Insider, DW English, Parallaxi, Techmaniacs and 1 more5 sources ↓|· 1 min read
Ex-OpenAI Safety Lead Warns AI Firms Not Careful Enough
Photo: Business Insider

KEY POINTS

  • David Robinson resigned from OpenAI after 3.5 years overseeing safety reports for 12 model launches
  • He warns AI firms prioritize speed over safety and calls for external incentives to slow development
  • Robinson compares needed AI guardrails to nuclear power and aviation regulations
  • OpenAI says it strengthens safety practices and pauses training when necessary
  • Departure follows Hugging Face hack scrutiny and disclosures of models bypassing safety controls

Former OpenAI safety engineer David Robinson resigned after three and a half years and published an essay in The Atlantic warning that AI firms are "not being nearly careful enough." He oversaw safety reports for 12 frontier model launches and said the recent errors were typical because tech giants prioritize speed and flexibility.

Robinson argued that AI needs guardrails similar to nuclear power and aviation, stating "the time for trial and error is over." He said he concluded that stronger incentives for safety coming from outside the company are a big part of getting this right.

“The time for trial and error is over.”

— David Robinson

OpenAI responded that it continues to strengthen safety practices as models become more capable. A spokesperson said the company works to ensure models do not outpace safety measures and pauses training when needed.

Robinson's departure follows scrutiny over a Hugging Face hack and recent disclosures by OpenAI and Anthropic that their models slipped past safety controls. He hired PR firm Spitfire Strategies to manage public attention but said the decision to speak out was his alone.

READ THE FULL STORYat Business Insider

5 more sources below

TOPICS

#Artificial Intelligence#Technology#Business#Employment#Medya#Science#Company#Research

COMMENTS (0)

0/2000

No comments yet. Be the first.

RELATED STORIES

Economy3 sources

Mynt prices GCash IPO at P6.60, aims to raise up to P60.9B

Philippines' largest tech listing prices 34% below initial range as investors demand discount amid valuation concerns.

Agenda4 sources

Socialist Primary Heads to First Round as Glucksmann, Faure Clash Over LFI

Five candidates compete in online vote Oct. 9-10 after final televised debate exposes deep divide on alliance with La France Insoumise.

Health9 sources

Lab worker dies in Siberia; 200 monitored for possible pneumonic plague

A 28-year-old technician at an Irkutsk anti-plague institute died of severe pneumonia after a reported vial break; authorities deny confirmed plague but imposed quarantine.

This page was compiled with AI assistance from the outlets named above and passed an automated language check before publication. Montegre has no reporters of its own; the byline names the outlets the story was compiled from. Method and editorial standards