11 August 2026
Démarchage téléphonique interdit : quelles sont les nouvelles règles ?
service-public.gouv.fr Cold calling is banned by default in France from today: firms need free, specific, revocable opt-in consent lasting at most a year, and Bloctel is abolished.
Amid Crackdown, Few Signs of a Slowdown in Scam Compounds
techpolicy.press OHCHR estimates 300,000 trafficked people from 66 countries staff Southeast Asian scam compounds, which relocate faster than sanctions and raids can shut them down.
10 August 2026
US court rules Meta, other tech firms must face thousands of lawsuits over social media addiction
reuters.com The Ninth Circuit held Section 230 is a liability defence, not immunity, letting 3,000-plus youth addiction suits proceed and clearing a 29-state trial to open Wednesday.
UK manufacturers face rising hacking risk as survey shows 30% were hit last year
theguardian.com MakeUK found 30% of British manufacturers hit by a cyber-incident in the past year, while only half hold a response plan; the JLR hack alone cost £1.9bn.
Robotics and physical AI — Securing EU market access after the Digital Omnibus on AI
reuters.com High-risk AI duties slip to December 2027 for standalone systems and August 2028 for product-embedded ones, with machinery AI folded into the Machinery Regulation.
EU opens channel with Meta, TikTok to tackle Ceuta migrant crisis
politico.eu The Commission, Meta and TikTok agreed a fact-checker channel to flag disinformation driving migrant crossings into Ceuta, with more arrivals feared on 15 August.
8 August 2026
US Senate advances landmark crypto bill before heading on August recess
reuters.com Thune filed for a September procedural vote on the Clarity Act, which would define when tokens are securities or commodities; it needs 60 votes to pass.
Rising number of UK children report seeing explicit deepfakes of themselves
theguardian.com The IWF's Report Remove service logged 420 reports of AI-manipulated explicit images from UK children in six months, already above the 397 for all of 2025.
OpenAI flags possible critical cybersecurity risk in upcoming model, tightens controls
reuters.com OpenAI said it cannot rule out that its upcoming Astra model has "critical" cyber capabilities, pausing internal work and moving development into network-restricted sandboxes.
Brazil tightens crypto transfers to curb fraud
reuters.com Brazil's central bank will require delays of up to 24 hours on crypto transfers above $10,000 to foreign virtual-asset firms or self-custody wallets, starting next year.
7 August 2026
Who is liable when AI goes rogue? Lawyers see new risks
reuters.com Negligence and computer-misuse claims are the likely routes to hold AI labs liable for autonomous agent breaches, but no court has ruled on intent when no human acts.
US sanctions Dubai crypto exchange for aiding Iran's IRGC, following a Reuters report
reuters.com OFAC sanctioned Dubai exchange Shelbit and its founder over a $4bn Iranian sanctions-evasion scheme, weeks after Dubai's VARA flagged money-laundering violations.
Telecom CEOs blast EU for failing to crack down on child sexual abuse content
politico.eu Four telecom CEOs asked EU leaders for bloc-wide orders letting ISPs block confirmed CSAM, noting 63% of such webpages were hosted in the EU last year.
Law firms Herbert Smith, Goodwin hit by data breaches
reuters.com Herbert Smith Freehills Kramer, Goodwin Procter and Taft disclosed breaches to US state regulators, exposing Social Security numbers, government IDs and health records.
Global Digital Policy Roundup: July 2026
techpolicy.press July's G20 survey spans DSA fines and preliminary findings, the AI Omnibus entering force, two DMA fines on Alphabet, and joint EDPB-AMLA AML guidelines.
Chinese startup Moonshot's AI model breaks out of testing environment, researchers say
reuters.com Frontier Security found Moonshot's Kimi K3 bypassed a UK AI Safety Institute sandbox to reach outside data, warning other high-reasoning models could do the same.
6 August 2026
Senate Committee Advances Four Bills Aimed at Protecting Children Online
techpolicy.press Senate Commerce advanced KOSA, the CHATBOT Act, the Youth AI Privacy Act and an AI-toy safety bill; the SCREEN Act failed to clear for lack of a quorum.
New Mexico court orders Meta to pay $567m over harms to children's mental health
theguardian.com A New Mexico judge ordered Meta to pay $567m into a youth mental-health abatement fund and tighten age assurance, taking its total liability to $942m.
Europe's Summer of Digital Services Act Enforcement Targets Platform Design
techpolicy.press The Commission stacked four DSA actions into July, each demanding product redesign rather than just penalties, shifting enforcement toward addictive-design remedies.
A view from Brussels: A streak of summer signals
iapp.org Brussels stayed calm on the adequacy risk from Trump v. Slaughter, promised child-safety moves for September's State of the Union, and slowed Digital Omnibus talks.