Urgent.News

What's breaking now, across thousands of outlets.

AI

OpenAI says it didn’t fire 3 safety researchers for 'speaking out' on AI risks

OpenAI on Friday defended firing three AI safety researchers after they warned in an open letter that their dismissals would have a "chilling" effect.

Abstract editorial illustration

OpenAI has defended its decision to terminate three AI safety researchers, asserting that the dismissals were not in retaliation for speaking out about the risks associated with artificial intelligence. The three researchers, Jasmine Wang, Mikita Balesni, and Tomek Korbak, were fired following an internal investigation that uncovered a breach of policy regarding sensitive information handling.

OpenAI's CEO, Sam Altman, clarified that the firings were unrelated to the researchers' concerns about AI safety. Instead, the company stated that the researchers violated policies surrounding sensitive information, leading to their dismissal. In a separate open letter, the disgraced researchers voiced their concerns that their firings could create a chilling effect on OpenAI's culture, discouraging employees from expressing their thoughts and operating freely.

Despite the researchers' assertions, OpenAI maintained that the terminations were motivated by policy breaches and not by the researchers' attempts to raise safety concerns. The company also emphasized its commitment to preserving monitorability of frontier AI models, which it considers crucial to its ongoing research efforts. OpenAI expressed deep regret over the situation and praised the researchers' contributions to AI safety during their tenure.

Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at businessinsider.com →

More in AI

The AI Co-Pilot Dilemma: Does Code Generation Stifle Authentic Programming Practice?

AI coding assistants now write boilerplate, suggest entire functions, and explain error messages in seconds. For many developers, that speed feels like a gift. For others, it raises an uncomfortable question: does code generation stifle authentic programming practice?

  • AI coding assistants generate code rapidly, raising concerns about authentic practice.
  • Removing friction from programming may hinder learning if solutions aren't understood.
  • Beginners should write their own versions of code to develop deeper comprehension.

How to test whether coding agents discover the skills you need

A coding skill can be clear, correct, and still fail to help if an agent never finds it. Testing a skill by naming it directly only answers whether the agent can use it after it has been loaded. It does not show whether an agent will discover it from the kind of request a teammate or customer actually makes.

  • Conduct task-based assessments for coding agents to gauge skill discovery.
  • Analyze execution traces to measure skill loading without explicit guidance.
  • Track three metrics: skill activation, utilization, and application performance.

More from Friday 9 October →