THE RUMP REPORT

9 TOP AI TOOLS VULNERABLE TO 'HALLUSQUATTING' ATTACK THAT BUILDS MASSIVE BOTNETS...

ars-technica · JUL 8

FILING: CHINESE AI MAKER Z.AI SEEKS $4B SHARE SALE AFTER STOCK ROCKETS 1,500% SINCE JANUARY...

techmeme · JUL 8

REPORT: CHINESE CYBER WATCHDOG CLAIMS CLAUDE CODE SECRETLY BEAMS LOCATION, IDENTITY DATA OFFSITE

techmeme · JUL 8

SAMBANOVA VALUATION ROCKETS TO $11B... MONTHS AFTER INTEL'S $1.6B TAKEOVER TALK

techcrunch-ai · JUL 8

REPORT: MICROSOFT CUTS AI SPENDING... SHIFTS WEIGHT TO ITS OWN MODELS

techcrunch-ai · JUL 7

OPENAI'S TOP FUTURIST OUT AFTER NEARLY NINE YEARS ON AI SAFETY

wired-ai · JUL 7

STARTUP LAUNCHES APP TODAY TO STOP AI SCAMMERS POSING AS KIDNAPPERS DEMANDING RANSOM...

techcrunch-ai · JUL 7

REPORT: US STARTUP'S SELF-DRIVING WAR MACHINES COMPLETE 1,100+ MISSIONS IN UKRAINE

techcrunch-ai · JUL 7

REPORT: AMAZON'S NEXT-GEN ALEXA 'MOONRAKER' TO TACKLE COMPLEX, MULTISTEP TASKS... PRICE TAG: $100M+ IN GPUS

techmeme · JUL 8

LAWMAKERS WEIGH CRACKDOWN ON US FIRMS' CHINESE AI ADDICTION... HOUSE PROBE EYES RISKS

techmeme · JUL 8

DISCORD ADMITS: AI MOD BOT WRONGFULLY BANNED USERS OVER HARMLESS IMAGES SINCE MAY

techcrunch-ai · JUL 7

REPORT: USERS ALREADY PUSHING BACK AS META LETS ANYONE TAG THEIR INSTAGRAM INTO AI-GENERATED IMAGES

verge-ai · JUL 7

REPORT: ALTMAN IN TALKS WITH TRUMP TO HAND U.S. GOVERNMENT A 5% STAKE IN OPENAI

mit-tech-review · JUL 7

SOURCES: MINIMAX BUILDING 2.7 TRILLION-PARAMETER AI MODEL... LARGEST CHINESE AI MODEL CURRENTLY IN DEVELOPMENT

techmeme · JUL 8

[PAPER] STUDY: AI TRAINED TO FOOL ITS OWN JUDGE... PASS RATE HITS 94% AS TRUE ACCURACY STUCK AT 20%

arxiv · JUL 7

[PAPER] STUDY: AI REASONING MODELS CAN "THINK" THEIR WAY INTO WRONG ANSWERS... DUBBED "THINKING-INDUCED HALLUCINATION"

arxiv · JUL 7

[PAPER] STUDY: NEW AI MODEL IS ONLY SYSTEM TO PASS SIMULATED RADIOLOGY BOARD EXAM

arxiv · JUL 7

[PAPER] STUDY: AI IMAGE FILTERS BREAK DOWN WHEN THE RULES CHANGE... SAME PICTURE, DIFFERENT VERDICT

arxiv · JUL 7

[PAPER] STUDY: HIDDEN CODE WATERMARK HITS 100% DETECTION, ZERO FALSE POSITIVES ON CLEAN FILES

arxiv · JUL 7

[PAPER] STUDY: SCIENTISTS CRACK OPEN SELF-DRIVING AI'S BLACK BOX... THEN REACH IN AND FIX ITS BAD DECISIONS BY HAND

arxiv · JUL 7

[PAPER] STUDY: RESEARCHERS EXPOSE FIRST-EVER ATTACK ON INFRARED AI DEPLOYED IN 'SECURITY-CRITICAL' SETTINGS

arxiv · JUL 7

[PAPER] STUDY: AI CODE AGENT PROVES 4,257 SOFTWARE LEMMAS IN A ROW... ZERO FAILURES, NO HUMAN HELP

arxiv · JUL 7

[PAPER] STUDY: SOME AI "UNLEARNING" METHODS DON'T ACTUALLY ERASE YOUR DATA... AUDIT EXPOSES WHICH ONES FAIL

arxiv · JUL 7

[PAPER] REPORT: AI'S HOTTEST BUZZWORD HAS NO DEFINITION... EVERYONE'S BUILDING ONE ANYWAY

arxiv · JUL 7

[PAPER] REPORT: RESEARCHERS BUILD FRAMEWORK FOR AUTONOMOUS CONTROL OF FIBER-OPTIC NETWORKS... VALIDATED IN TESTBED

arxiv · JUL 7

[PAPER] NEW AI GUARDRAIL TRAINED TO REASON... THEN ORDERED TO STAY SILENT AT RUNTIME

arxiv · JUL 7