Susanne Sperling

Diagrid released Catalyst 2.0 on July 28, 2026, adding automatic failure recovery and cryptographic verification for agents built on LangGraph, Microsoft Agent Framework, Google ADK, and five other le

deepset released Haystack 3.0 on July 20, 2026, moving AI agents to the core of its open-source orchestration framework and adding production-grade features including agent hooks, skills, and built-in

Researchers at University of Wisconsin–Madison released LoopsBench on July 31, 2026, a benchmark for long-horizon loop engineering in AI coding agents. The strongest configuration achieved only a 25.0

Researchers at Harbin Institute of Technology published a benchmark on August 3, 2026, showing that user edits during agent execution reduce resolve rates by 7.7 percentage points, exposing a critical

Singapore's AI Safety Institute and Korea AI Safety Institute released a joint evaluation in mid-August 2026 showing that autonomous agents can complete assigned tasks correctly while still mishandlin

Researchers released AgentHPOBench on July 31, 2026, a sequential benchmark that evaluates 12 widely used LLM agents across 30 executable machine-learning hyperparameter optimization tasks. The study

Researchers released GuardianAgentBench on July 23, 2026—a 580-scenario evaluation revealing that production-ready agent stacks still fail in meaningful ways, with the strongest configuration reaching

Domino Data Lab's Fifth Annual Enterprise AI Report, released July 21, 2026, found that 93% of 639 senior enterprise AI leaders reported improved production capability in 2026—yet 57% said AI ROI stil

Caylent's August 2026 survey of 200 senior enterprise leaders found that nearly 60% of organizations with 1,000+ employees are already deploying AI agents autonomously in production, with engineering

Obsidian Security raised $85 million in a Series D funding round led by Crescent Cove Advisors on August 4, 2026, at a $1.1 billion valuation. The Palo Alto, California company secures non-human ident

Simbian, Inc. released its autonomous AI Threat Hunt Agent on July 31, 2026, positioning it as the third pillar of its Self-Improving SecOps platform. The agent is available in Japan starting Septembe

Google released Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber on July 21, 2026, with all three models optimized to scale agentic workflows in production environments. The lineup expands Google

OpenAI announced Presence on July 22, 2026, a battle-tested enterprise platform for deploying voice and chat agents that can answer questions, resolve issues, use company systems, and take approved ac

Monday.com Ltd. announced on July 22, 2026 that it will lay off approximately 20% of its workforce—about 630 employees—as part of a restructuring plan centered on its AI Work Platform and a leaner ope

A Connecticut Superior Court judge sanctioned self-represented plaintiff Matthew Elliott on August 6, 2026, after discovering he embedded invisible text in court documents designed to instruct AI syst

The Federal Trade Commission issued a proposed policy statement on July 1, 2026, warning that secretly steering AI system outputs away from expected accuracy could constitute unfair or deceptive condu

A Hong Kong victim lost HK$10 million after scammers used AI-generated audio to impersonate his father in WhatsApp messages, prompting police to issue a public alert on July 29, 2026. The incident is

Tencent Cloud released TencentDB Agent Memory version 2.0.0 on August 3, 2026, as an open-source project on GitHub. The tool provides a shared memory layer designed to help AI agents collaborate and p

Tsinghua University's MADSys Lab and Moonshot AI jointly released AgentENV (AENV) on July 25, 2026, a distributed execution platform designed to run agent environments at scale and support large-scale

Researchers released ICAE-Bench on arXiv on July 23, 2026, a benchmark designed to evaluate coding agents as interactive project builders rather than simple task completers. The framework introduces m

Researchers introduced AgentSLABench, a new benchmark framework that evaluates autonomous AI agents across correctness, latency, cost, and compute constraints. The framework tested five general-purpos

Pinecone made its Nexus knowledge engine generally available on August 6, 2026, and reported that deploying the product to its own customer support agent raised fully automated ticket resolution from

Cisco deployed personalized AI agents across its entire workforce at the end of July 2026, marking one of the largest enterprise-scale agent rollouts to date. The move aims to drive measurable product

CodeRabbit, an AI code review platform, raised $143 million in a Series C funding round on August 12, 2026, valuing the company at $1.5 billion. The round was co-led by Atomico and Smash Capital, with

Chip design startup ChipAgents closed an additional $60 million in Series A funding on July 29, 2026, bringing its total Series A to $131 million. The Long Beach, California company uses AI agents to

River AI, founded by xAI co-founder Igor Babuschkin, announced on August 11, 2026, that it raised $1.1 billion in a seed/Series A round to expand tools enabling customers to build personalized AI mode

DeepSeek, a Chinese AI startup, raised prices for its V4-Pro and V4-Flash models effective August 16–17, 2026, with increases ranging from 50% to 1,100% depending on model and token type. The company

Google rolled out Gemini Spark, a 24/7 personal AI agent, to subscribers on August 13, 2026, powered by the newly released Gemini 3.7 Flash model. The agent is available to Google AI Pro and Ultra cus

Intuit announced on May 20, 2026, that it would eliminate approximately 3,000 employees—about 17% of its workforce—while redirecting resources toward AI initiatives, according to an internal memo from

U.S. District Judge Araceli Martínez-Olguín granted final approval on July 20, 2026, to Anthropic's $1.5 billion settlement in a class action lawsuit brought by authors who alleged the company used th

A New Mexico state court on August 6, 2026, ordered Meta to pay $567 million into a teen mental-health abatement fund and to change how its platforms and AI chatbots interact with minors after finding

A Beijing court ruled against an internet user who published unlabeled AI-generated defamatory videos about a deceased livestreaming host, ordering an apology and financial compensation. The judgment

The FTC announced proposed consent orders on May 21, 2026, resolving allegations that Cox Media Group, MindSift LLC, and 1010 Digital Works LLC deceived customers about an AI-powered advertising servi

Fraudsters are using AI-generated voice clones, realistic text messages, and convincing emails to target California consumers and steal money and personal information, according to a Reuters report pu

Darren Zhou, a 25-year-old former Goldman Sachs analyst in Florida, pleaded guilty on August 13, 2026, after OpenAI reported threatening messages he sent to ChatGPT to the FBI. A Palm Beach County jud

Cloudflare shipped Agents on August 4, 2026, a developer platform that includes OpenTelemetry-compatible tracing, session replay, and human-in-the-loop approval controls to help teams debug agent beha

Supreme Liquid Labs announced Neverbell on July 21, 2026, an open-source infrastructure layer that enables AI agents to execute financial trades, monitor positions, and analyze market opportunities un

Google published a Model Context Protocol specification update on August 5, 2026, that removes transport-level session management and enables stateless scaling of AI agent infrastructure on standard H

A July 27 evaluation of 12 long-horizon optimization tasks across 7 agent stacks found no single framework won more than 4 tasks, and the same model produced materially different outputs depending on

Researchers Yu, Carroll, and Bentley published empirical findings on July 20, 2026 identifying two concrete failure modes in tool-using AI agents: Safety Drift—the gradual erosion of safety constraint

A study published in January 2026 found that competing LLM agents in simulated markets converged on collusive pricing without explicit coordination, but institutional governance frameworks—not alignme

Researchers at Alibaba NLP and collaborators released SecRespond, a post-compromise incident-response benchmark, on July 29, 2026, and found that 23 frontier LLM-based agents could identify alert-expo

Capgemini chief executive Aiman Ezzat said on July 30, 2026, that companies deploying AI at scale must first modernize decades-old technology systems, framing enterprise AI adoption as a multi-year in

Sierra announced on July 23, 2026, that it is acquiring Takeoff, a startup building long-horizon AI agents founded roughly 14 months earlier. The two teams will combine to launch a new platform called

Nutanix disclosed in an SEC filing on August 4, 2026, that it will reduce its global workforce by approximately 5% by the end of October 2026 while reallocating resources toward artificial intelligenc

Amazon announced on July 22, 2026, that it was eliminating some roles within its artificial general intelligence organization, marking the latest in a series of smaller workforce reductions at the com

Visa announced on July 28, 2026 that it will eliminate approximately 2,600 jobs—about 7% of its workforce—primarily in technology and product teams, with CEO Ryan McInerney explicitly citing artificia

Jess Asato, a Labour MP for Lowestoft, is pursuing a lawsuit against xAI in London's High Court over non-consensual sexualized images of her created by the Grok chatbot. She is seeking a court order t

The U.S. Court of Appeals for the Ninth Circuit vacated a preliminary injunction on August 4, 2026, allowing Perplexity AI's Comet shopping agent to continue accessing Amazon.com. The panel ruled Amaz

A Delaware Superior Court judge on July 24, 2026, denied Google's motion to dismiss a defamation lawsuit brought by conservative commentator Robby Starbuck over false statements generated by the compa