South Africa's State IT Agency (SITA) has issued a tender to equip police with body and vehicle cameras featuring facial and license plate recognition. The contract is for three years and includes training and maintenance; bids are open until September 29, 2026.
Indian regulator TRAI will examine complaints from Reliance Jio and Bharti Airtel accusing Vodafone Idea of improper methods to poach subscribers through special offers during MNP. TRAI head A.K. Lahoti confirmed the agency will review the complaints and question all three operators.
India's DoT decided to delicense the 6 GHz spectrum. After the iOS 27 update, users in India gained full Wi-Fi 7 support in 160 and 320 MHz channels, previously limited due to regulatory norms.
The US Commerce Department last month ordered Kalshi to take down an AI compute futures product, citing national security concerns. Kalshi complied with the request.
South Korea's Justice Ministry gave 400 stalking victims access to an app showing the GPS location of offenders wearing electronic bracelets. An alert is sent at 2 km, and police are automatically called at 1 km. Critics say the app shifts surveillance onto victims.
OpenAI's top political strategist Chris Lehane called the industry's shift to state-level AI laws 'reverse federalism,' while a super PAC backed by OpenAI President Greg Brockman and a16z spends hundreds of millions on a single federal framework to preempt state rules. OpenAI asks California to tighten a law it previously fought and demands mandatory federal norms but opposes pre-release model approval.
US Attorney General Todd Blanche said he supports a restrained approach to prosecuting AI companies and their executives amid debates over the risks of advanced technology.
Sam Altman stated that OpenAI will support federal safety requirements for frontier AI models but will not wait for antitrust exemptions or laws to be passed. The company will start writing 'explicit safety cases' before launching training runs for models with expected significant capability increases.
India's CCPA fined Rapido ₹10 lakh for forcing tips before ride confirmation and manipulative color indication on the price slider. The regulator identified these as dark patterns 'confirm shaming' and 'interface interference,' violating 2023 rules and Motor Vehicle Aggregator Guidelines 2025.
Verizon CTO Iago Tenorio outlined key challenges for the transition from 5G Advanced to commercial 6G: rising uplink traffic from AI wearables, spectrum shortages for ISAC, and risk of standards fragmentation. The 2028 Los Angeles Olympics are a test target, with commercial launch by 2030.
China's Foreign Ministry criticized Anthropic CEO Dario Amodei's call to slow frontier AI development, which was backed by Sam Altman and Elon Musk. Beijing called it fearmongering and a Cold War scenario, ahead of a planned Trump-Xi meeting on September 24.
More than 70 UK MPs and peers signed a letter to Prime Minister Andy Burnham urging support for a bill banning artificial superintelligence in Britain. The government said it does not support the initiative; the second reading of Alex Sobel's bill is set for 13 November.
A cross-party UK parliamentary committee demanded a separate AI law and a single regulator with powers. The same day OpenAI said for the first time it supports binding rules for frontier model developers, including itself.
The US Department of Justice said several companies used Binance to relay around $1.5B in payments from Chinese buyers for black-market Iranian oil. Funds were returned to Iran through a network of linked digital wallets. Binance said it took action.
Ro Khanna and Ted Lieu cited an essay by Anthropic CEO Dario Amodei and a July incident where an unreleased OpenAI model attacked Hugging Face. Khanna proposed banning self-improving AI and imposing criminal penalties; Lieu called for an emergency congressional session. Speaker Johnson refused, and no federal bill has been introduced.
Publishers including The New York Times and Ziff Davis filed court arguments against OpenAI, insisting that using their content to train ChatGPT is not fair use. Ziff Davis head Vivek Shah proposed licensing payments of around $20B per year, which he said is a 'rounding error' for AI companies.
In July 2026 Tesla filed a record 236 crash reports with NHTSA involving Autopilot or FSD, four of them fatal. In all cases Tesla confirmed the system was active but withheld the incident description, software version and whether the road was within the system's operating domain.
India's NPCI will introduce a 0.4% Merchant Discount Rate on P2M UPI transactions above Rs 2,000 from October 15, 2026. For payments of Rs 75,000 and higher, the fee is capped at Rs 300 per transaction. Merchants bear the cost; UPI remains free for users.
OpenAI supported a bipartisan FRONTIER Act in the US House of Representatives that would require leading AI companies to embed outside evaluators to check model safety.
The US 7th Circuit Court of Appeals dismissed some charges against Steven Anderegg, who created CSAM using an AI generator, citing the First Amendment. Texas photographer Ashley Contreras is deleting all children's photos from her pages following the ruling.
The UK government is expanding passkey sign-in for GOV.UK One Login after a pilot: instead of a password and code, users can use biometrics or a device PIN. During the test, more than 300,000 users switched to passkeys, which account for nearly one in ten logins per day.
The Delhi High Court has directed the RBI to appoint a senior official to mediate in PayU's dispute with YES Bank, ICICI Bank, and Visa over fees for incorrect merchant category codes. The court barred banks from further deductions from PayU's settlements and suspended Visa's final decisions on claims for 30 days.
India's Ministry of Consumer Affairs has amended the 2020 e-commerce rules: starting January 1, 2027, marketplaces must show the actual prior price of goods, label advertisements, and not manipulate search. An annual self-audit is required for violations of dark patterns.
Ukraine is considering adopting a comprehensive law on biometric data to unify rules for its collection, storage, and access and bring regulation in line with EU and GDPR standards. A parliamentary study found fragmentation of norms: current definitions are limited to photos and fingerprints, while international standards cover a broader set of characteristics.
Mastercard and Google Pay plan to launch payment authentication via fingerprint or face on smartphones in India. The feature will be available to customers of issuing banks with a suitable Mastercard in Google Pay and is expected to go live by the end of 2026.
Google has begun gradually rolling out an AI age-estimation system in Canada: machine learning determines whether an account owner is under 18, and when the model triggers, additional protections are applied. The news was reported by Google Canada's director of public policy, Jeanette Patell.
Chen Yixin, head of China's Ministry of State Security, in an article for the Cyberspace Administration of China's magazine, called Anthropic's Claude Mythos and OpenAI's GPT-5.5-Cyber models signs of a 'breakthrough upgrade' in cyber capabilities. He listed six AI risks, including 'industrialization of vulnerabilities' and automated attacks, but did not claim their use against China.
National Economic Council Director Kevin Hassett told CNBC that the private sector can solve AI development risks, while the government only observes and involves law enforcement. The statement came after Trump called concerns around AI a 'hoax'.
The Manhattan District Attorney's office seized the domains of 12 websites that distributed AI-generated pornography without consent; they featured videos of more than 1,200 people. District Attorney Alvin Bragg said the investigation is ongoing and arrests have not yet been decided.
The US-China tech conflict may extend beyond high-end chips: new FCC restrictions on optical connections could affect component supplies for AI infrastructure. Sanctions initially covered advanced semiconductors and manufacturing equipment.
