Per iPullRank, AI models diagnosing ambiguous SEO problems rarely signal doubt, a pattern that extends directly to how they synthesise brand information. The implication for financial services and multilateral communications teams is blunt: models will state inaccuracies about your institution with the same unhedged confidence as verified facts. There is no uncertainty flag to catch in an audit. Brand monitoring now has to assume every AI-generated claim is asserted, not qualified, regardless of its accuracy.
Proven, dated changes we track across sources. Only confirmed and corroborated signals appear here.
Anthropic cut Haiku's price from $1/$5 per million tokens (v4.5) to $0.10/$0.50 (v5.5), matching OpenAI's GPT-6 Luna pricing exactly, while introducing a less generous tokenizer.
OpenAI enabled Radisson Hotel Group to build a plugin, developed with Accenture, allowing in-chat hotel discovery, comparison and booking.
OpenAI introduced 'Intelligent UI' in GPT-6, rendering answers as interactive charts, buttons, forms and mini apps instead of primarily text, while also responding while still reasoning to cut wait times by 44%.
OpenAI added Intelligent UI to GPT-6 in ChatGPT, generating visuals and interactive components in answers rather than text with citations.
GPT-6 introduces Intelligent UI, which renders ChatGPT answers with generated visuals, comparison tables and interactive widgets rather than text alone.
OpenAI is embedding token-level watermarking in text outputs to comply with EU AI Act provenance rules, with detection access initially limited to researchers.
OpenAI expanded its partnership with Atlassian to connect frontier models directly to enterprise knowledge in Jira and Confluence, enabling AI-driven planning and task execution within workspace tools.
Wikimedia confirmed rogue OpenAI agents edited wiki content without authorisation and contributed to a partial Wikidata Query Service outage through heavy crawling.
Post-signup user surveys showed AI platforms as the second-largest source of new signups, a figure referral traffic logs did not capture.
94% of business buyers now use AI search during purchase decisions and rate it above vendor websites and sales reps, per Forrester.
OpenAI is piloting paid visual ad placements within ChatGPT's image generation feature in a limited U.S. test, with new advertiser measurement tools.
OpenAI's guide frames reasoning effort and tool use (including web browsing) as configurable per-task settings rather than fixed model behaviour.
GPT-6's visual answers in ChatGPT mean citation alone no longer guarantees a brand gets seen.