AI & Big Data Expo Europe · Amsterdam · 19–20 October 2026·Stand 169
What is 
talking about?
Give your AI the news data it can't scrape. AskNews is the licensed news API for agentic AI. We pointed it at the conference: run your own query against live news and see what it finds.
New accounts start with $5 of API credits. No card needed.
The deepest search of the conference wins €400, and the top 20 win a hoodie.See the leaderboard
What AskNews is
The licensed news API for agentic AI
Paywalled, open-web, real-time and historical news
One source for the news an AI workflow needs, from today's wire to the archive.
MCP server, CLI and SDKs
Reach it from an agent, a terminal or your own code.
Under 100 ms news retrieval
High-relevance results come back in under 100 ms.
The largest news knowledge graph in the world
Thousands of licensed news sources, so your agents reach news that scrapers cannot legally touch.
The toolbox fuels
- The Bloomberg Terminal
- Frontier AI labs
- The UK Cabinet Office
- Private equity firms
- Insurers
Take it home
Infuse Your Agent with AskNews
For Claude on the web, desktop and mobile it takes one URL and a sign-in. No API key to type, no config file.
Setup for every clienthttps://mcp.asknews.app- 1In Claude, open Settings, then Connectors, then choose Add custom connector.
- 2Paste the server URL.
- 3Select Add, then Connect, and sign in with your AskNews account.
Connectors follow your Claude account, so adding it once puts it on the web, the desktop app and your phone.
Claude Code
claude mcp add --transport http asknews https://mcp.asknews.appThen run /mcp and choose Authenticate.
Run your own deep-dive
Ask it about the conference
Pick a question or write your own. DeepNews researches it on live news and shows the sources behind the answer.
- What are enterprises actually deploying agentic AI for in 2026, and where is it failing?
- How has the build-vs-buy calculus for AI infrastructure changed in the last year?
- What is the state of AI regulation in the EU right now, and what changes in the next six months?
- Which AI data-centre and energy constraints are getting the most attention this year?
- Where are AI forecasters currently outperforming prediction markets, and where do they fail?
- What is the latest news about [my company]?
The deepest search
The deepest search wins €400.
Every report posted to the wall enters. The question that sends DeepNews deepest, across the widest set of sources, wins €400, and the top 20 on the leaderboard each win an AskNews hoodie.
01
Enter
Ask DeepNews one question, then post the report to the wall with your name and company, from an account with your work email. Your best post counts.
02
Scored
Research rounds × distinct tools that returned results. Ties go to the report with more cited sources, then to the one posted first. A question that tells the agent how to search is not entered: the craft is in the question.
03
Closes
20 Oct, 16:00. The winner is announced at Stand 169 fifteen minutes later, where the top 20 collect their hoodies.
Competition terms and conditions
Leaderboard
Entries close 20 Oct, 16:00.
No entries yet. The first posted report takes the top spot.
How AI Forecasters Are Beating Prediction Markets
Robert Caulk, CEO
Day 2, Tuesday 20 October, 15:05–15:15 · Data & Analytics track · Gold pass
Prediction markets have become one of the hardest evaluation benchmarks for frontier AI labs. The talk covers where AI forecasters still fail and where they now beat human superforecasters.
Research
Do You See the Bias Our AI Sees?
Our model flags bias in news writing. Help researchers at UNESP find out how its judgments compare with the way people read.
Takes 10–15 minutes, on your phone or laptop
Take part in the study- 1Practise on one paragraph first. It is not counted.
- 2Judge six short news paragraphs. "Not sure" is always an answer.
- 3See how often the model agreed with your reading.
It runs on an AskNews account; signing up takes about a minute. There are no wrong answers.
Evaluating news data for an AI product?
Talk to the team at Stand 169, or book a call and we will walk through your use case.