NeuralStackAI Law Radar’s 29 July 2026 changelog shows a fast-moving AI rulebook that is still being rewritten entry by entry. The latest update adds several new obligations and one new U.S. state bill, while also tightening confidence levels on existing records where primary sources now confirm the details.
One of the biggest additions is the UK’s Data (Use and Access) Act 2025 automated decision-making reform, which the tracker says came into force on 5 February 2026. The register says it replaces the UK GDPR’s Article 22 default prohibition with notification, representation, human-review and contest duties.
The same update also adds the UK’s DUAA 2025 deepfake offence under section 138, which the tracker says took effect on 6 February 2026. AI Law Radar says the offence criminalises creating or requesting a non-consensual intimate deepfake even where it is never shared, and it distinguishes that rule from separate offences aimed at tool suppliers.
In the United States, the tracker adds Illinois SB 343, an algorithmic rental price-coordination ban that amends the Illinois Antitrust Act. The bill has passed the state legislature and is on Governor Pritzker’s desk with a 29 August 2026 action deadline, so AI Law Radar is tracking it as proposed and without an effective date.
The changelog also revises the way it categorises several price-coordination bans. It says the New Jersey FAIR Act, Maryland’s Protection From Predatory Pricing Act and Illinois SB 343 now sit under Prohibited AI practices because each bans a use of an algorithm outright.
China’s AI Agents Implementation Opinions, issued by the CAC, NDRC and MIIT on 8 May 2026, are also added to the register. The tracker does not spell out the full practical scope in the changelog entry, but it marks the opinions as a new item in its coverage.
Other changes in the same batch are mostly corrections and confidence adjustments rather than substantive shifts. AI Law Radar says the UK AI (Regulation) Bill [HL] remains at first reading in the House of Lords with no progress since introduction, and that evidence from the UK government’s Blueprint for AI Regulation continues to support the view that the bill is unlikely to advance as drafted.
Why this matters
The pattern in these updates is that AI regulation is no longer limited to broad framework laws or frontier-model oversight. The newest items target automated decisions, intimate deepfakes and algorithmic price coordination, which suggests regulators are moving toward specific harms rather than waiting for a single comprehensive AI statute.
That shift matters for companies because the compliance question is becoming more sectoral and more conduct-specific. Firms that thought AI governance meant only model documentation or transparency disclosures now have to watch antitrust, criminal law and consumer-protection style rules as well.
It also matters because the tracker is giving more weight to primary-source confirmation. Several entries were upgraded in confidence after official parliamentary, legislative or government materials confirmed their status, which indicates the underlying rules are becoming clearer even as the policy landscape keeps changing.
For the near term, the most important signal is that the register keeps growing across jurisdictions rather than converging on one model. The UK, U.S. states and China are all adding different kinds of AI obligations, and the practical burden for deployers is likely to keep shifting as more of these rules move from proposal to force.