Top 10 AI Stories Today — 17 August 2026

1. IMF Warns Agentic AI Set to Fundamentally Reshape Global Payments

IMF Notes · Finance & Banking

The IMF has published a new note warning that agentic AI — autonomous systems capable of independently managing and executing complex tasks — is poised to transform payments by cutting transaction costs, improving liquidity, and reducing fraud, but also introducing new governance risks. The report highlights that removing human oversight from payment flows could accelerate capital circulation but may also create systemic vulnerabilities if AI agents act without adequate controls. The findings carry direct implications for European and UK banks already navigating the EU AI Act’s newly enforceable high-risk AI rules.

https://www.elibrary.imf.org/view/journals/068/2026/004/article-A001-en.xml

2. AI Fuels Record Surge in Self-Represented Litigants Flooding Federal Courts

ABA Journal · Law & Legal

Artificial intelligence has driven a dramatic rise in pro se (self-represented) federal court filings, with the share of such cases jumping from a steady 11% to 16.8% in fiscal year 2025 — a historic shift tied directly to powerful AI tools like ChatGPT. Research shows that roughly 18% of federal civil complaints now contain AI-generated text, and some state bar officials attribute at least 30% of a spike in disciplinary complaints to AI use. While AI is boosting access to justice for those who can’t afford lawyers, courts and attorneys warn the flood of AI-assisted filings is creating significant new burdens and accuracy concerns.

https://www.abajournal.com/web/article/Tell-me-how-to-get-even-with-my-lawyer

3. AI Drug Discovery Market Set to Triple to $25 Billion by 2035

GlobeNewswire / ResearchAndMarkets · Healthcare & Pharmaceutical

A new industry report released on August 13, 2026 projects the global AI in drug discovery market will grow from $8.6 billion in 2026 to approximately $25 billion by 2035, a compound annual growth rate of 12.6%. The growth is being driven by continued investment, expanding datasets, accelerating patent activity, and demand for cost-efficient drug development. As AI platforms mature and regulatory guidance evolves, AI is expected to become increasingly central to pharmaceutical research and pipeline development worldwide.

https://www.globenewswire.com/news-release/2026/08/13/3344697/28124/en/ai-in-drug-discovery-market-report-2026-3rd-edition-trends-and-forecast-to-2035.html

4. French Press Body Files Complaint Over Google AI Summaries Hurting Publishers

Rio Times Online · Media & Marketing

France’s press publishers association APIG has filed a formal complaint against Google, arguing that its AI-generated search summaries repeat a pattern of bypassing fair negotiation with news publishers — a pattern Google was already fined €500 million for in 2021. The complaint is part of a broader European push, with Italy’s communications authority also referring the matter to the European Commission. Publishers argue AI summaries divert users away from original content, threatening smaller and independent outlets.

https://www.riotimesonline.com/french-press-body-google-ai-summaries-2026/

5. UK Signs £2bn AI Deal to Train 60,000 Soldiers Annually

National Security News · Defense & Security

The UK Ministry of Defence has signed a £2 billion contract led by Raytheon UK to train up to 60,000 British soldiers per year using artificial intelligence, as part of the Army’s Collective Training Transformation Programme. The deal aims to give troops greater battlefield realism and supports the government’s ambition to make the British Army significantly more lethal by 2035. It is expected to create around 400 jobs across Britain, including 270 new roles in AI, software engineering, and cloud computing.

https://nationalsecuritynews.com/2026/07/british-army-to-train-60000-soldiers-a-year-using-ai-in-2bn-deal/

6. AI Infrastructure Boom Shifts from Chip Race to Supply-Chain Race

InvestorIdeas / GlobalData · Supply Chain

A new GlobalData report published on 5 August 2026 finds that AI infrastructure spending is now being constrained not by chip demand, but by shortages in memory, advanced-node fabrication capacity, and electrical power. Semiconductor lead times reached 40 weeks in March 2026, with memory ICs among the most acutely constrained components as AI data centres consume up to 70% of all memory chips produced globally. Intel’s 18A manufacturing node reached high-volume production in January 2026 with improving yields, attracting foundry interest from Nvidia, Microsoft, Amazon, and Apple.

https://www.investorideas.com/news/2026/technology/08052-ai-infrastructure-chip-supply-chain-globaldata.asp