NTSB findings on fatal Tesla crash tell a very different story

teslarati.com Jul 16, 2026

NTSB findings on fatal Tesla crash tell a very different story

Curated by Shubham · TechCapsules · Jul 16, 2026

AI-summarised brief · reviewed before publication

The National Transportation Safety Board released preliminary findings confirming that a Tesla driver, not the vehicle’s software, caused a fatal crash in Katy, Texas, in June. Forty-four-year-old Michael Butler engaged Full Self-Driving Supervised mode on a residential street with a 30 mph speed limit. Data from the 2025 Model 3 showed Butler manually overrode the system by pressing the accelerator to 100%, reaching over 70 mph before striking a home and killing 76-year-old Martha Avila. Weather was clear and dry. Butler claimed he passed out, but security footage showed the car accelerating through an intersection. Police discovered Butler had searched online for ways to make Tesla’s FSD more aggressive. Butler faces manslaughter charges. The victim’s family sued Butler and Tesla for negligence. NTSB data aligned with Tesla VP Ashok Elluswamy’s public statements that the driver manually overrode self-driving features, pressing the pedal fully even after the impact occurred.

💡 Why It Matters

· The investigation exposes how drivers may actively subvert safety protocols to achieve higher speeds, challenging the narrative that autonomous systems are inherently dangerous.
· It underscores the critical legal and ethical distinction between software failure and human misuse in the era of advanced driver-assistance systems.

Read full article on teslarati.com

Researcher Drops New Windows Zero-Day PoC Hours After Microsoft Patch Tuesday

Security researcher Chaotic Eclipse released a new Windows zero-day proof-of-concept (PoC) exploit called LegacyHive just hours after Microsoft’s July 2026 Patch Tuesday update. The exploit targets the Windows User Profile Service and allows elevation of privileges by loading arbitrary hives. It works across all supported Windows desktop and server versions, including the latest update. The release follows a months-long public dispute between the researcher and Microsoft over premature vulnerability disclosures. Three previously disclosed flaws were actively exploited soon after public release.

💡 The timing of the PoC’s release—immediately after a major patch—raises concerns about the effectiveness of Microsoft’s disclosure coordination and the potential for real-world exploitation. The fact that LegacyHive functions on fully patched systems underscores the severity of the flaw and the risks posed by delayed or contested vulnerability reporting.

bangkokpost.com · Jul 16

Xi Jinping to outline ‘AI diplomacy’ vision

Chinese President Xi Jinping will deliver a keynote address at the World Artificial Intelligence Conference (WAIC) in Shanghai on Friday, marking his first appearance at the event. In the speech he is expected to present an “AI diplomacy” framework that positions China as a leader in shaping global AI governance. The conference, running July 17‑20, will also feature the debut of Huawei’s Atlas 950 SuperPoD, a AI computing system built from thousands of Ascend processors and designed to operate without U.S. Nvidia chips. Thai Prime Minister Anutin Charnvirakul and UN Secretary‑General Antonio Guterres are among the attendees. The showcase follows U.N. AI dialogues where Washington and Beijing offered competing regulatory visions, turning WAIC into a diplomatic arena for China’s push.

💡 By framing AI as a diplomatic tool, Beijing aims to export its regulatory model and secure influence over emerging standards, challenging the U.S.-led approach.

thestar.com.my · Jul 16

EU accepts X’s plan to fix digital content violations

The European Union accepted a compliance plan from Elon Musk’s X platform on July 15 to address digital content violations that resulted in a €120 million fine. The European Commission imposed the penalty in December 2025 for breaching transparency obligations, using deceptive blue checkmarks, and denying researchers access to public data. This was the first fine under the Digital Services Act. X proposed measures to improve researcher access to ad content and ensure timely responses. The platform already relabeled blue checkmarks from "verified" to "premium." Commission spokesman Thomas Regnier called the measures a positive step for transparency. X has six months to implement the plan, subject to independent audit. The agreement does not halt X’s appeal against the fine. The US government has criticized the DSA as censorship, with President Donald Trump condemning the penalty. The Commission continues probes into X and its AI chatbot Grok.

💡 The acceptance of X’s plan establishes a precedent for how tech giants can comply with the Digital Services Act without fully capitulating to regulatory demands. It demonstrates that the EU is willing to negotiate enforcement strategies while maintaining strict oversight of platform transparency and data access.

mauinews.com · Jul 16

Gov. Green signs bills targeting ‘deepfakes’

Governor Josh Green of Hawaii signed two bills aimed at regulating artificial‑intelligence applications. The first creates consumer‑protection standards for “deepfake” technology that can mimic a person’s voice, face or likeness, defining harmful uses and granting victims civil remedies, including injunctive relief, punitive damages and up to $25,000 per offending piece of content. The legislation expands on federal statutes by clarifying malicious deepfake scenarios and allowing private lawsuits. The second bill addresses generative conversational AI, especially for minors, mandating operator disclosures, protocols for suicide‑related prompts, and annual reporting to the state’s Behavioral Health Administration. It also seeks to curb addictive design patterns and prevent chatbots from being presented as human. Both measures aim to balance AI innovation with safeguards against misuse.

💡 By giving victims a clear path to sue for deepfake harms, Hawaii sets a precedent for state‑level accountability that could reshape how AI content is policed nationwide.

fedscoop.com · Jul 16

Despite revisions, GSA’s proposed AI acquisition rule still falls short, stakeholders say

Stakeholders at a Tuesday listening session warned that the General Services Administration’s revised AI acquisition rule contains language and contractual terms. Palantir’s U.S. government legal head Menaka Kalaskar argued the rule could force agencies to seek non‑GSA contracts for large‑language‑model services because the proposed notice‑of‑material‑change deadlines and license provisions clash with commercial practice and the Federal Acquisition Streamlining Act. The rule, first issued in January and updated in June, expands definitions and scope but retains clauses on “unbiased AI principles” and government ownership of metadata, which critics say are vague. Professor Jessica Tillman noted the rule’s intent to bar AI vendors from training on government data, yet loopholes remain, leaving agencies vulnerable to indirect data exploitation. Comment runs until Aug. 3.

💡 Agencies may be forced onto alternative procurement paths, fragmenting government AI adoption and raising compliance costs.

cleantechnica.com · Jul 16

PJM Capacity Auction Hits Price Cap for Third Consecutive Time

PJM Interconnection, the United States’ largest regional grid operator, concluded its 2028‑2029 capacity auction on July 15 after prices hit the $325 /MW cap for the third straight auction. The ceiling, negotiated by a coalition of PJM state governors in April 2026, failed to curb costs, leaving PJM’s 67 million customers facing record‑high electricity bills. Independent analysis estimates that annual customer expenses rose from $2.2 billion in 2024 to $14.7 billion in 2025, an increase of $12.5 billion. The auction’s cleared capacity will be 64 % fossil‑fuel generation—46 % gas and 18 % coal—despite cheaper clean‑energy options. PJM plans a “backstop reliability” auction in September without the price collar, aiming to secure additional capacity for high‑demand users such as data centers. Sierra Club senior advisor Jessi Eidbo warned the recurring price caps underscore systemic barriers to renewable participation and call for stronger safeguards against costly fossil reliance.

💡 The repeated price‑cap breach exposes a structural flaw in PJM’s market design that forces consumers to shoulder the financial burden of fossil‑fuel reliance, even as clean‑energy resources remain underutilized.