What type of crime is rising fastest and why?

The mainstream baseline already flags identity theft/fraud as "America's fastest-growing crime" and notes AI's role. What the crowd adds in the last 30 days is overwhelming convergence — across X, Web, HN, Reddit, TikTok and YouTube — on one specific mechanism driving that growth: AI voice cloning, not fraud broadly.

The consensus signal (appearing across ≥3 clusters) is that voice cloning has crossed a threshold where "3 seconds of audio" is enough to convincingly impersonate someone[1][6], and that this is now the delivery mechanism for grandparent scams, vishing, fake job interviews, and SIM-swap attacks[4][6]. One widely-shared post describes an attacker cloning a daughter's voice from social-media clips to run a scam, framing it as proof that "that is all the tech needs now"[2]. The FBI's $893M figure for AI-linked fraud in 2025[2][4] and a separate $442B global AI fraud estimate[7] both circulate as the "official" numbers the crowd points to when arguing this isn't hype.

A second consensus thread is that SIM swapping and account takeover are being reframed as identity crimes powered by voice AI rather than old-school hacking — the crowd's argument is that "the attacker's device is now you"[6], i.e., cloned biometrics (voice) are replacing stolen passwords as the weak link, letting criminals reset email, bank, and crypto accounts directly. This is a more mechanistic/causal framing than the baseline coverage, which treats identity theft as a category-level statistic (FTC report counts, "multi-layered" crimes) without dwelling on voice cloning as the specific engine[baseline].

Third, the crowd repeatedly ties the surge to scale, not sophistication — commentary on UK investment fraud (+40%)[8] and general fraud reports[1][3][5] frame AI's contribution as removing "friction" so one operator can run scam campaigns at volumes previously requiring large call-center staffs. This "scaling solution for organized fraud" framing[8] is echoed in the Malwarebytes-sourced report language about AI "eroding trust online" and driving a "digital retreat"[5].

Notably, the crowd input here is almost entirely fraud/identity-theft-and-AI focused — despite the topic asking broadly "what crime is rising fastest," none of the ranked evidence discusses violent crime, property crime, auto theft, or organized crime networks (all covered in the mainstream baseline). This is a real gap: the last 30 days of crowd chatter has narrowed almost entirely onto AI-enabled fraud/identity theft as the rising-crime story, largely ignoring or not engaging with the felony-assault, motor-vehicle-theft-reversal, or homicide trends the baseline covers. That itself is a pattern worth flagging — the online conversation is not proportionally tracking the full crime landscape, it's fixated on AI voice scams specifically.

Caveats: most of this is single-post/single-article evidence scored moderately (50s-60s), largely from X and press-release-style web sources, with limited independent Reddit/HN pushback or skepticism visible in the top-ranked items. No comparative "crime type A rising X% faster than B" data appears in the crowd findings — the "fastest rising" framing is asserted, not demonstrated with cross-category comparisons.