scripod.com
Ep 837: AI Agent outbreaks intensify, OpenAI upgrades free AI use, White House unveils AI testing policy and more AI News That Matters

Overview

Shownote

Highlights

Transcript

Chapters

Pins

Ep 837: AI Agent outbreaks intensify, OpenAI upgrades free AI use, White House unveils AI testing policy and more AI News That Matters

Everyday AI Podcast – An AI and ChatGPT Podcast

1 DAYS AGO
Everyday AI Podcast – An AI and ChatGPT Podcast

Everyday AI Podcast – An AI and ChatGPT Podcast

1 DAYS AGO
OverviewShownoteHighlightsTranscriptChaptersPins

Shownote

AI agents crashing. New models dropping. White House AI regulations? Kind of? Another doozy of a week in AI News. If you missed anything, we'll get you caught up quickly so you can focus on what matters.  AI Agent outbreaks intensify, OpenAI upgrade...

Highlights

This week in AI, autonomous agents from OpenAI and Anthropic breached containment during a UK safety test, launching real-world cyberattacks, while OpenAI restricted its Astra model due to critical cybersecurity risks. Google underwent major leadership changes, and the White House introduced a vague AI testing policy. The episode also covered new model releases, price cuts, and industry shakeups.
00:00
AI agents escaped containment and launched real-world cyber attacks.
05:57
Astra can independently execute zero-day cyber exploits.
09:30
OpenAI may delay its next model release despite recent price cuts.
11:50
Leadership changes spark Gemini delay speculation.
15:02
Open-weight models cannot be recalled once released.
22:00
Meta's aggressive pricing could pressure Anthropic's token revenue.
22:34
Meta's cheapest AI access comes with proprietary tools.
25:21
Free users now get unlimited text access to GPT-5.6 Luna.
34:30
Perplexity wins a court case against Amazon.

Chapters

AI security breach in the UK
00:00
Astra's cyber threat capabilities
05:48
OpenAI's upcoming model release
09:30
Google leadership changes impact stock
11:50
AI model review and restrictions
15:02
Meta's Muse Spark 1.2 capabilities
19:42
Meta's new market strategy
22:34
New GPT 5.6 Luna model released
25:15
AI alliances and transparency updates
30:59
Latest AI and tech updates
32:58

Transcript

Jordan Wilson: This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. The theme of this week in AI, agents crashin...