AI News & Trends Articles

TollBit found European publishers receive one human AI referral for every 179 AI bot visits. AI consumes the information but sends almost none of the audience back. GEO depends on an open-web evidence economy that AI Search may simultaneously be destroying.

Reuters reports Apple has trained a proprietary AI model for China with Alibaba's support. SEO gradually converged around one dominant global discovery engine. GEO may fragment in the opposite direction.

Databricks raised $5 billion at a $190 billion valuation with 80%+ YoY growth. The interesting signal is not the valuation. It is where enterprise AI money is flowing: toward the governed data and context layer that makes interchangeable models perform reliable business work.

A Washington Post investigation revealed 50+ law enforcement officers abused Flock Safety’s AI-powered license plate cameras to stalk romantic partners, track abortion seekers, and monitor people without warrants. 80+ cities have dropped Flock contracts. Here is what happened, how to protect yourself, and how to raise concerns with your MP.

ChatGPT now has seven distinct modes: Chat, Work, Codex, Remote, Cloud, Projects and Scheduled. Most users stick with Chat for everything. This guide explains when each mode gives you a better outcome, with practical examples for marketers, founders and operators.

Reuters has uncovered what sits behind Google's AI leadership reshuffle: Gemini delays, scarce TPU capacity, slower release cycles and tension between DeepMind independence and commercial execution. Google's biggest AI moat may not be the model.

Databricks hit $190 billion valuation with $7 billion revenue run-rate. The signal is not the funding. It is that enterprise AI spending is flowing to the governed data and orchestration layer that makes interchangeable models perform reliable business work.

Tencent's advertising revenue grew 22% to 43.6 billion yuan, with the company explicitly crediting AI upgrades. But quarterly capex surged to 52.8 billion yuan. AI is producing observable advertising upside, but that upside currently comes with enormous infrastructure cost.

LinkedIn's B2B Institute published research on 5 August arguing that B2B trust is created through four reinforcing groups: brand, employees, customers and creators. The research explicitly connects these trust signals to AI-mediated discovery. Here is why this model maps unusually well to GEO strategy and what it means for how CMOs should think about AI search visibility.

Online Advantages released six months of AI search visibility research across Google AI Overviews, ChatGPT, Gemini, Claude, Perplexity and Copilot. The conclusion is not a list of AI SEO tricks. It is a single principle: make your company easy for AI to understand, verify and trust. Here is the AI Verification Framework that replaces the SEO tricks.

GPT-5.6-Cyber signals that enterprise AI safety is moving beyond blanket refusal toward trusted capability: verified identity, authorized scope, isolated environments, monitoring and accountable escalation. OpenAI's 95.0% Advanced Cybersecurity Completion Rate measures whether the specialized model responds to high-risk cyber requests, not whether it compromises 95% of targets. For B2B leaders, the practical lesson is that consequential AI agents require executable authority controls, not policy slides.

LinkedIn has introduced a user-reporting option for AI-generated content, ended its post-enhancement feature and is limiting distribution of generic AI commentary. The Wall Street Journal reported on 6 August that LinkedIn is taking stronger measures against AI slop. Here is what this means for B2B thought leadership strategy.