NeuralStackAI Law Radar’s latest change log shows a notable reversal in Illinois: SB 343 is no longer being tracked as an AI rental-price-coordination ban because the AI language was removed before the bill was transmitted to Governor Pritzker. The register says a Senate floor amendment had added the ban, but a later House committee amendment replaced that content entirely with unrelated eminent-domain provisions. The bill sent to the governor on June 30 contains no AI provisions, and AI Law Radar says the confidence level is now high that no AI obligation exists under that bill number.
The update also tightens the status of Rhode Island’s therapy AI services ban, H 7349. AI Law Radar says the effective date is now confirmed as the signing date, June 22, and the tracker has removed earlier language that treated the date as unconfirmed. The underlying takeaway is straightforward: the rule is already in force, and the register no longer treats that timing as an open question.
At the federalized European level, the tracker marks a major milestone for the EU AI Act. Article 50, covering chatbot transparency and synthetic-media disclosure, and Article 101, covering GPAI fining powers, both entered into force on August 2, 2026. AI Law Radar’s entry treats those obligations as no longer pending, but active.
The California AI Transparency Act also moved into force on the same date, according to the changelog. AI Law Radar says SB 942, as amended by AB 853, now requires GenAI providers with more than 1 million monthly users to offer an AI-detection tool and content provenance disclosures. The tracker also adds AB 853 as a new entry, with Phase 1 now in force and Phase 2 for large online platforms and device manufacturers set to activate on January 1, 2027.
Beyond those headline changes, the register adds several coverage gaps and corrects earlier entries. It now includes the UK Data (Use and Access) Act 2025 automated decision-making reform, the UK deepfake offences created by DUAA section 138, and China’s AI Agents Implementation Opinions from the CAC, NDRC, and MIIT. It also says the UK AI (Regulation) Bill remains stalled at first reading in the House of Lords and is unlikely to advance as drafted.
Why this matters
For AI companies, these updates underscore how compliance cannot be built around static bill summaries alone. A proposal can look live one week and disappear from the final text before signature, as Illinois SB 343 now shows, which means legal teams need amendment-level tracking rather than relying on bill titles or early summaries.
The same logic applies in Europe and California, where transparency obligations are no longer theoretical. With the EU’s chatbot disclosure and synthetic-media rules in force, and California’s provenance and detection duties now active, product teams that deploy user-facing generative AI have to translate policy changes into interface, labeling, and detection workflows quickly.
AI Law Radar’s broader July and August revisions also point to a maturing regulatory landscape that is spreading across jurisdictions and issue areas. The tracker is no longer limited to high-level model governance; it now spans consumer disclosures, deepfake rules, automated decision-making, rental-pricing coordination, and criminal offenses tied to non-consensual synthetic media.
For vendors, platforms, and in-house counsel, the practical lesson is that AI regulation is increasingly defined by the exact text in force on a given day, not by the direction of travel alone. That makes revision histories, amendment trails, and effective-date checks as important as the statutes themselves.