AI News — Monday, August 17, 2026

10
Stripe Reportedly Acquires AI Gateway Startup OpenRouter for $7B+

Stripe is reportedly acquiring OpenRouter, an AI gateway startup, for over $7 billion, signaling a major consolidation in the AI infrastructure market.

TechCrunchindustry
9
Salesforce Rolls Out New Slackbot AI Agent to Compete in Workplace AI

Salesforce has launched a new Slackbot AI agent, intensifying its competition with Microsoft and Google in the rapidly expanding workplace AI market.

VentureBeatproduct
8
Full-bandwidth transformer: A Novel Architecture for Enhanced AI Performance

A new transformer architecture is proposed that utilizes full-bandwidth processing, potentially leading to significant improvements in AI model performance.

Hugging Faceresearch
8
How We Got an LLM to Draw Charts Without Ever Touching a Pixel

This article details a novel method for enabling Large Language Models to generate charts purely through symbolic representation, bypassing direct pixel manipulation.

Dev.toproduct
8
Anthropic CEO: AI Backlash is 'Fundamentally a Crisis of Trust'

Anthropic's CEO attributes the growing AI backlash to a fundamental crisis of trust, emphasizing the importance of ethical considerations and safety in AI development.

TechCrunchindustry
7
Listen Labs Raises $69M After Viral Stunt to Scale AI Customer Interviews

Listen Labs secured $69 million in funding following a successful viral marketing campaign, aiming to expand its AI-powered customer interview platform.

VentureBeatindustry
7
Why People Aren’t Buying Mark Zuckerberg’s AI Future

This article explores the reasons behind public skepticism and the lack of widespread adoption for Mark Zuckerberg's ambitious vision of an AI-driven future.

TechCrunchindustry
7
Self-Geometry: GT-Free and Plug-and-Play Test-Time Adaptation for 3D Vision Models

Researchers propose Self-Geometry, a novel test-time adaptation method for 3D vision models that ensures geometric consistency without requiring ground truth data.

Hugging Faceresearch
7
An AI4AI Framework for Visual Token Pruning to Optimize Visual Models

Researchers introduce an AI-driven framework designed to prune visual tokens, aiming to significantly optimize the efficiency and performance of visual models.

Hugging Faceresearch
7
H2R-Bench: Benchmarking Human-to-Robot Manipulation Video Generation in World Models

A new benchmark, H2R-Bench, is introduced to rigorously evaluate the generation of human-to-robot manipulation videos within advanced world models.

Hugging Faceresearch
7
Thought-Level Beam Search for Enhanced Reasoning in AI Models

A new 'thought-level beam search' method is introduced to significantly improve reasoning capabilities in AI models by exploring multiple, more complex reasoning paths.

Hugging Faceresearch
6
How RingCentral Builds AI-Native Workflows from Engineering to Operations

RingCentral shares insights into its comprehensive strategy for integrating AI natively across its entire operations, from core engineering to operational processes.

OpenAI Blogindustry
6
What Building an AI-Native Finance Function Taught Me

An individual shares practical lessons learned and insights gained from the experience of establishing an AI-native finance function within an organization.

OpenAI Blogindustry
6
AVA-Encoder: Towards Agent-Native Video Representation Learning

AVA-Encoder is presented as a significant step towards learning video representations that are inherently suitable and optimized for AI agents.

Hugging Faceresearch
6
Maglev: Introducing Sliding Recurrent Memory for Sequential Models

Maglev introduces a novel sliding recurrent memory mechanism designed to enhance the efficiency and context handling capabilities of sequential AI models.

Hugging Faceresearch