The 'poison AI' movement wants to corrupt ChatGPT and Gemini to make them useless — but it comes with a huge risk of collateral damage
THE SO WHAT
Organized attempts to poison training data turn the open web into a contested surface—labs aren't the only ones at risk, any model trained on public data inherits that fragility. Enterprises relying on external data sources should be investing in provenance, filtering, and synthetic or first-party corpora rather than assuming "more internet" equals better models.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIAI Storage Demand Still Growing, Says WD CEO
AI buildout is turning storage into a structural bottleneck, not a commodity line item—hyperscalers are effectively pre-buying capacity while equity markets punish any near-term guidance wobble. If you're planning AI-scale data products, lock in your storage roadmap and vendor terms now rather than assuming spot capacity will be there at the right price.
Applied AIOpenAI updates the default model for free users to GPT-5.6 Luna, adds unlimited text chats for free users, rolls out an improved GPT-5.6 Sol version, and more
Putting GPT-5.6 Luna in front of free users with unlimited chats collapses the capability gap between free and paid and normalizes frontier-level assistance as the baseline. If your product assumes "we're safer because users are on weaker free models," that assumption just expired.
Applied AIOpenAI introduces Agent Plugins, an open standard for bundling skills and MCP servers, and says its steering committee includes Amazon, Microsoft, and Cursor
An "Agent Plugins" standard with Amazon, Microsoft, and Cursor on the steering committee is a bid to formalize how skills and MCP servers plug into agent ecosystems—this is middleware, not just a feature. If you're building agents or tools, you now have to decide whether to align with this standard early or risk fragmentation and rework later.
Applied AIOpenAI is giving ChatGPT free users unlimited text chats
Unlimited text chats on free and Go tiers turns ChatGPT into a default utility layer for a much larger user base—usage, telemetry, and funnel, not short-term ARPU, are the priority. If your product relies on rate limits to differentiate or throttle AI usage, assume that moat just shrank.