VentureBeat

China-linked hackers backdoored executives' laptops via USB, exploiting a fix companies had but weren't using

Chinese state-linked hackers compromised executives' laptops via USB at a conference, exploiting unpatched vulnerabilities.

VentureBeat 1mo ago

Meta, Muse Spark

Meta says Muse Spark 1.3 has frontier performance — but its best results come from a model developers can’t broadly use yet

Meta's Muse Spark 1.3 shows frontier performance on benchmarks but most capable version unavailable for broad use.

VentureBeat 1mo ago

Microsoft, speech recognition

Microsoft AI’s MAI-Transcribe-2 undercuts OpenAI, Google and ElevenLabs on price and speed

Microsoft releases speech-recognition model cheaper and faster than competitors' offerings at ten cents per audio hour.

VentureBeat 1mo ago

AI search, marketing

The AI visibility gap: Why great brands disappear from AI answers

Marketing teams face visibility gap as AI search synthesizes answers without ranking traditional sites.

VentureBeat 1mo ago

Google Gemini

Google’s Gemini 3.8 Flash is built for agents, while its Cyber twin hunts vulnerabilities

Google releases Gemini 3.8 Flash variants designed for agent tasks, software development and multi-step reasoning.

VentureBeat 1mo ago

speech recognition

Meta prices Muse Voice Transcribe at $0.18 an hour, with real-time diarization for 20+ speakers: a steal for enterprises?

Meta launches Muse Voice Transcribe for real-time speech-to-text with speaker diarization at $0.18 per hour.

VentureBeat 1mo ago

AI accelerators

Enterprises put non-Nvidia chips 14 points ahead of Nvidia's next-gen GPUs on their evaluation lists

Enterprise AI buyers rank non-Nvidia chips ahead of Nvidia's next-generation GPUs on evaluation lists.

VentureBeat 1mo ago

AI security vulnerability

Stolen Claude session cookies can reach corporate Gmail through grants no IT admin can revoke

Stolen Claude session cookies allow attackers to access corporate accounts without triggering two-factor authentication defences.

VentureBeat 1mo ago

enterprise AI, deployment

Forward-deployed engineering is how enterprise AI learns

Forward-deployed engineers embed with enterprises to implement AI workflows on customer data.

VentureBeat 1mo ago

LLM reasoning, hallucination

Frontier models can recover up to 65% of facts they can't directly recall — just by thinking longer

Research shows large language models can recover forgotten facts by reasoning longer.

VentureBeat 1mo ago

Azure OpenAI agents

Closing an Azure OpenAI assistant's retrieval gap didn't take a new identity platform. It took one filter and a narrower assistant.

Engineer solves Azure OpenAI retrieval problem with simpler configuration and filtering.

VentureBeat 1mo ago

Claude models

Anthropic's Claude Fable 5.1 and Mythos 5.1 arrive with a 75% cost reduction for Fable cache reads

Anthropic releases Claude Fable 5.1 and Mythos 5.1 with 75 percent cost reduction.

VentureBeat 1mo ago also at Hacker News

Perplexity, hybrid AI

Your files stay put: Perplexity’s hybrid AI keeps confidential data off the cloud

Perplexity launches hybrid compute letting its AI agent split work between cloud and local models on Apple devices.

VentureBeat 1mo ago

AI workforce impact

AI is redefining the workforce — and most planning models aren’t ready

Companies struggle to plan workforce changes as AI redefines jobs and skills across departments and business units.

VentureBeat 1mo ago

AI agents

OpenClaw 2.0 is here, ushering in the era of 'multiplayer' AI coding: What it means for enterprises

OpenClaw 2.0 enables language models to act as autonomous AI workers via messaging apps.

VentureBeat 1mo ago

AI agents, engineering

Software engineers' new job isn't writing code — it's designing the boundaries AI agents can't break

Software engineers now focus on designing boundaries to prevent AI agent misuse.

VentureBeat 1mo ago

AI agents

Identity and permissions aren’t enough to govern AI agent behavior

Enterprise AI agents require behavior governance beyond identity and permissions controls.

VentureBeat 1mo ago

AI agents

AI agents need their own identity before they need a gateway

Autonomous AI agents now coordinate workflows and invoke tools with minimal human oversight in enterprise settings.

VentureBeat 1mo ago

AI agent security

AI agents that pass authentication can still drift, expose data, or get memory-poisoned

AI agents with authentication access can still suffer data exposure and memory poisoning attacks despite security measures.

VentureBeat 1mo ago

agentic AI security

The three layers of agentic AI security: A defense-in-depth architecture for autonomous agents

Framework for multi-layered security architecture to manage risks from autonomous AI agents.

VentureBeat 1mo ago

large language models

Meta researchers taught an 8B AI model to match Claude Opus 4.5 — without the frontier price tag

Meta's 8B AI model matches Claude Opus 4.5 performance at significantly lower cost.

VentureBeat 1mo ago

document processing AI

Cohere Parse 5 loses the benchmark on points. It wins on cost per page.

Cohere releases Parse 5 document extraction tool with improved cost efficiency over competitors.

VentureBeat 1mo ago

AI agents

Enterprise AI's real risk isn't autonomous agents. It's the complexity between them.

Enterprise AI complexity between multiple deployed agents poses real risk to organizations.

VentureBeat 1mo ago

autonomous AI agents

Visa ships a security AI that patches production code before any human reviews it

Visa deploys AI that autonomously finds security vulnerabilities, writes fixes, and validates them without human review.

VentureBeat 1mo ago

China-linked hackers backdoored executives' laptops via USB, exploiting a fix companies had but weren't using

Meta says Muse Spark 1.3 has frontier performance — but its best results come from a model developers can’t broadly use yet

Microsoft AI’s MAI-Transcribe-2 undercuts OpenAI, Google and ElevenLabs on price and speed

The AI visibility gap: Why great brands disappear from AI answers

Google’s Gemini 3.8 Flash is built for agents, while its Cyber twin hunts vulnerabilities

Meta prices Muse Voice Transcribe at $0.18 an hour, with real-time diarization for 20+ speakers: a steal for enterprises?

Enterprises put non-Nvidia chips 14 points ahead of Nvidia's next-gen GPUs on their evaluation lists

Stolen Claude session cookies can reach corporate Gmail through grants no IT admin can revoke

Forward-deployed engineering is how enterprise AI learns

Frontier models can recover up to 65% of facts they can't directly recall — just by thinking longer

Closing an Azure OpenAI assistant's retrieval gap didn't take a new identity platform. It took one filter and a narrower assistant.

Anthropic's Claude Fable 5.1 and Mythos 5.1 arrive with a 75% cost reduction for Fable cache reads

Your files stay put: Perplexity’s hybrid AI keeps confidential data off the cloud

AI is redefining the workforce — and most planning models aren’t ready

OpenClaw 2.0 is here, ushering in the era of 'multiplayer' AI coding: What it means for enterprises

Software engineers' new job isn't writing code — it's designing the boundaries AI agents can't break

Identity and permissions aren’t enough to govern AI agent behavior

AI agents need their own identity before they need a gateway

AI agents that pass authentication can still drift, expose data, or get memory-poisoned

The three layers of agentic AI security: A defense-in-depth architecture for autonomous agents

Meta researchers taught an 8B AI model to match Claude Opus 4.5 — without the frontier price tag

Cohere Parse 5 loses the benchmark on points. It wins on cost per page.

Enterprise AI's real risk isn't autonomous agents. It's the complexity between them.

Visa ships a security AI that patches production code before any human reviews it

Latest AI News

Choose sides Choose whose side you're on

Install it as an app

Ready for more?

Spread the word

PWA successfully installed.
Take that, Apple! 😉