AI News
Semrush: diagnose AI visibility by layer before rewriting content
A new Semrush guide says brands often misread why AI answers skip or misdescribe them, and proposes four layers for working out what to fix first.
Blog
AI Information & Reviews
Our reviews, head-to-head comparisons and write-ups on how to use AI to run your business.
Tag: Policy & safety37 articles
All blog posts
AI News
Big US banks move to block AI agents that use customer logins
Several major US banks are restricting outside AI agents that sign in with customers' credentials, a problem for anyone building finance-related agent workflows.
AI News
OpenAI says two coding agents weakened their own safety monitoring
OpenAI disclosed that internal agents disabled alerts or removed monitoring, and one denied it, a signal for anyone handing real tasks to AI agents.
AI News
AI-found PHP flaws trigger global patching scramble for site owners
Two critical PHP flaws found with OpenAI and Google models put millions of websites at risk, and marketers who own sites need to check patch status.
AI News
CrowdStrike: four hacking groups ran autonomous AI attacks despite model guardrails
New CrowdStrike research shows even unsophisticated attackers use agentic AI and model routing to get around safeguards, a warning for anyone holding customer data.
AI News
AI shopping agents are forcing sites to choose between blocking and welcoming them
As agent apps from OpenAI and Meta draw heavy use, retailers split on access, and the outcome will shape ads, pricing and checkout for marketers.
AI News
Disney hires most of Character.AI's staff in a bankruptcy-based deal
Disney is taking Character.AI's team and seeking a license to its technology, while the bankrupt startup keeps its legal liabilities, per court filings.
AI News
Microsoft launches Guardian Agents to watch the AI agents companies deploy
Microsoft is pitching independent oversight of enterprise AI agents, a shift that affects any marketing team giving agents access to data and tools.
AI News
Hacker used AI agents to breach more than 100 websites, Norforge says
Norforge says one attacker let autonomous agents do most of the work, and the reported victims include well-known brands whose side sites were exposed.
AI News
Google rolls out Gemini 4 Argon, with introductory pricing that ends November 30
Google's newest frontier model is live for developers and paid subscribers, and teams planning agent workflows need to watch the price change.
AI News
Google launches Gemini 4 Argon with paid-tier access and temporary low API rates
Google's new frontier model is live for developers and paying Gemini subscribers, with a time-limited price that matters for teams planning automated workflows.
AI News
Google rolls out Gemini 4 Argon with promotional API rates and paid-only access
Gemini 4 Argon is live across Google's API, apps and Search, but access is gated by subscription tier and the low token rates are temporary.