
Google launches Gemini 4 Neon with Opus 5.5-level coding performance at 25% lower API token prices venturebeat.com/ai/google-la…

Obsessed with covering transformative technology.

Google launches Gemini 4 Neon with Opus 5.5-level coding performance at 25% lower API token prices venturebeat.com/ai/google-la…

Speaker announcement: Liran Zvibel, CEO & Co-Founder of @WEKA, joins us Dec 10 in Menlo Park. His question for the room: agents need working memory at every step, so where should the context layer live? Request an invite: bit.ly/4taAg6q

DIVD detected an AI-run attack in 19 minutes. Mounting its first response took 19 hours. Timing each stage separately shows where a response stalls. venturebeat.com/security/ai-…

It took UK government testers half as many queries to jailbreak pre-release GPT-6.2 Astra as GPT-6 Astra. Their tests also caught monitors slipping at OpenAI, Anthropic and Google. venturebeat.com/ai/openais-p…

#ai is making it easier for attackers to clone legitimate mobile apps, tamper with their #code, and make malicious versions look convincing. Here’s what #security teams need to know: #genai #cybersecurity #aiagents bit.ly/4Av6dsl

#mcp standardized how #ai agents discover and call tools. It never standardized how you prove who’s calling or what they’re allowed to touch. #genai #generativeai #aiagents #llms bit.ly/4C4LJZc


Instacart’s CTO stopped his engineers from reading code. 97% of their code is now AI-generated, and their shopping agent costs 1/8 what it did 8 months ago. On Beyond the Pilot, @Instacart’s CTO explains why the harness beats the model. Watch here: youtube.com/watch?v=ifZuCTsS…

"We'll switch it in a couple of days." Uday Kiran Medisetty on changing the model behind Uber's AI code review, which Uber tests on years of past code changes. venturebeat.com/infrastructu…

Vero launches free AI security community as enterprises confront the work of governing agents venturebeat.com/ai/vero-laun…

'Essentially unusable': OpenAI launches GPT-6.2 Astra and Sol to negative reception from AI power users venturebeat.com/ai/essential…

Attack agents built on ARTEX are raiding exposed AI gateways for API keys, then running on the keys they steal. The FBI counts more than 1,600 exposed ARTEX consoles. venturebeat.com/ai/amazon-sa…

PayPal keeps its business knowledge outside the model. When a better model comes out, "I'll just plug in and play," says PayPal's Ren Zhang. venturebeat.com/infrastructu…

CrowdStrike’s Adam Meyers says his team used to tell people they had 30 days to patch. Now it’s more like 30 hours. venturebeat.com/security/ai-…

Counterfeit mobile apps can steal #sso credentials, intercept #mfa tokens, and expose sensitive corporate #data. #security teams need to treat mobile devices as critical endpoints. bit.ly/4Av6dsl

Treat every #mcp server as an OAuth Resource Server and keep tokens narrow and short-lived. Don’t cache credentials at startup, or assume Dynamic Client Registration is a safe long-term bet. #ai #genai #generativeai bit.ly/4C4LJZc