Consultar a IA antes de uma decisão pessoal importante vs. decidir sem ela
Se você agir
Consultar a IA antes de decidir
22%
Se você não agir
Decidir sem aporte da IA
28%
Porcentagem de quem mais tarde se arrepende de cada escolha. As barras e o registro completo aparecem abaixo.
Estilo de vida
Última revisão 2026-05-11
Qualidade das evidências 4.0/5
Pontuação de revisão em oito dimensões segundo a
grelha de qualidade
. Cada dimensão pontuada de 1 a 5.
D1 Verificação das fontes
4/5
D2 Autoridade e independência das fontes
5/5
D3 Precisão da taxa de arrependimento
2/5
D4 Comparabilidade das fontes
2/5
D5 Padrão de Gilovich
4/5
D6 Qualidade da prosa
5/5
D7 Completude das ressalvas
5/5
D8 Qualidade da amostra
5/5
Média4.0/5
Dados proxy — não existe nenhuma pesquisa direta sobre arrependimento para esta decisão. As taxas são derivadas de pontuações de satisfação e dados de barreiras de acesso em vez de perguntas que perguntavam diretamente sobre arrependimento. Veja advertências abaixo.
Arrependimento por ação
Consultar a IA antes de decidir
22%
~22% dos que consultaram IA para uma decisão significativa relatam um resultado prejudicial ou problemático (proxy; 11% receberam recomendações de saúde inseguras; 21-49% das respostas médicas de IA são avaliadas como problemáticas em estudos controlados)
Adultos dos EUA que consultaram chatbots de IA para decisões de saúde, financeiras ou de vida
outubro-dezembro de 2025
Arrependimento por omissão
Decidir sem aporte da IA
28%
~28% dos que pularam a consulta à IA podem ter perdido informações úteis que poderiam ter melhorado sua decisão (proxy; 46-59% dos usuários de IA em saúde relatam benefícios concretos; 66-90% dos usuários financeiros de IA acharam valioso)
Adultos dos EUA que tomaram decisões significativas sem consultar IA
2025-2026
% se arrependem desta escolha
Consultar a IA antes de decidirDecidir sem aporte da IA
22%28%
balanced — Mais ou menos equilibrado — ambas as opções implicam arrependimento semelhante.
Decisões relacionadas
Decisões semanticamente semelhantes — mesmo terreno, compromissos diferentes.
Fazer um aumento de mamaNão fazer um aumento de mama
7,0%5,0%
Equilibrado
Aproximadamente equilibrado
About 1 in 4 US adults has now consulted AI for health information or advice, according to a West Health/Gallup panel of 5,660 adults surveyed in late 2025. Among those users, 46% felt more confident asking their providers questions afterward, and 59% used AI to prepare before a doctor visit — concrete stated benefits that non-users forgo. But the same survey found that 11% of AI health users reported receiving unsafe recommendations, and a parallel UCLA/BMJ Open study rating 250 AI responses to medical questions found 49.6% were problematic to some degree — mostly delivered with confidence and few caveats, making them difficult for users to identify as unreliable. An MIT Media Lab study published in NEJM AI documented that participants systematically overestimated AI medical reliability and could not distinguish AI-generated from physician responses, even when the AI response was inaccurate.
The regret arithmetic here is genuinely ambiguous, which is unusual in this dataset. The action-side risk (22% proxy, bounded by 11% unsafe-recommendation rate and 49.6% problematic-response rate) and the inaction-side opportunity cost (28% proxy, based on AI users’ reported benefits) are close enough that the entry is classified as ‘mixed’ rather than clearly inaction-dominates. How AI is used matters more than whether it is used: supplementing a scheduled doctor visit with AI research before attending is a different action category than using AI as a triage replacement for a symptom that warrants evaluation. The former has low action-risk and meaningful information benefit; the latter has higher action-risk and may cause harmful delay. The 14 million Americans who skipped a provider visit based on AI advice represent the higher-risk end of the use spectrum, though some of those skipped visits may have been genuinely unnecessary.
The honest summary of this entry’s evidentiary state: the AI consultation decision is too domain-specific, use-case-dependent, and rapidly evolving to generate a stable regret-pair estimate. Financial AI consultation self-reports are overwhelmingly positive (Wells Fargo: ~90% found results worthwhile), which would pull the inaction-regret figure up substantially if the financial domain were weighted equally with health. Medical AI consultation carries real documented risk of inaccurate confident advice in a domain where acting on wrong information has direct health consequences. The mixed classification reflects both the genuine uncertainty and the heterogeneity of “consulting AI” as a decision — something between a research tool and an advisor, with properties of each and the disclaimers of neither.
Fontes: acção
Registro de evidências
Cada número abaixo é o que cada fonte relatou, com a citação literal em que nos baseamos e como chegamos ao nosso valor. Clique em qualquer link para verificar diretamente.
2/3 fontes verificadas de forma independente palavra por palavra em relação à fonte citada
[1]West Health / Gallup — Millions of Americans Now Consult AI Before, After, and Sometimes Instead of Seeing a Doctor
Verificado
Fonte de referência
About 1 in 4 US adults (over 66 million) report having used AI tools or chatbots for health information or advice; 11% of US adults who used AI for health information reported receiving unsafe recommendations; 14% (~14 million) skipped a provider visit based on AI advice
Trecho
“"One in four U.S. adults — the equivalent of over 66 million Americans — report having used artificial intelligence tools or chatbots for physical or mental healthcare information or advice. ... 11% of AI health users reported receiving unsafe health recommendations from AI. 14% of recent AI health users skipped a provider visit based on AI advice. 46% felt more confident asking providers questions after using AI."
”
Dados da fonte de
2026-01-01
Acessado
2026-05-11
Verificação
Trecho obtido novamente de forma independente e confirmado palavra por palavra em relação à fonte citada durante nossa auditoria de fundamentação.
Cálculo
West Health / Gallup panel survey of n=5,660 US adults, fielded October–December 2025, margin of error ±2.1pp. This is the most methodologically rigorous large-sample US survey on AI health consultation outcomes. The "one in four US adults (over 66 million)" overall-prevalence figure grounds the entry's opening body sentence, which previously cited this survey for that figure without a matching excerpt/statistic. The 11% unsafe-recommendation rate is the primary action-regret proxy: it represents the share of users who received advice that was later identified as unsafe, which is a necessary precursor to regret even if regret is not directly measured. The 14% who skipped a provider visit based on AI advice is a behavioral-consequence measure; it does not directly map to regret (some may have correctly assessed that no visit was needed). The 22% action-regret estimate is bounded by the 11% unsafe-recommendation rate (lower bound) and the broader 21–49% problematic-response rates from academic accuracy studies (upper bound), centered at approximately 22% to reflect that problematic responses do not always produce regrettable outcomes.
[2]CIDRAP (University of Minnesota) reporting on UCLA / BMJ Open study — AI Chatbots Provide Poor Answers to Medical Questions Half the Time
Verificado
Fonte de referência
49.6% of AI responses to medical questions were rated as problematic in a blinded evaluation of 5 major chatbots (30% somewhat, 19.6% highly problematic)
Trecho
“"49.6% of AI chatbot responses to medical questions were rated as problematic — 30% 'somewhat problematic' and 19.6% 'highly problematic.' Chatbot responses were consistently given with confidence and certainty, with few caveats or disclaimers."
”
Dados da fonte de
2025-02-01
Acessado
2026-05-11
Verificação
Trecho obtido novamente de forma independente e confirmado palavra por palavra em relação à fonte citada durante nossa auditoria de fundamentação.
Cálculo
UCLA / BMJ Open study of 250 total questions across 5 major chatbots (ChatGPT, Gemini, DeepSeek, Meta AI, Grok), 10 questions each across 5 medical categories, data collected February 2025, published 2026. The 49.6% problematic-response rate establishes the upper bound of action-side risk for health-domain AI consultation. The key finding about confident framing with few caveats is the mechanism that converts problematic responses into potential harm: users cannot easily identify which responses are in the unreliable half. Used here to anchor the upper bound of the 22% action-regret estimate; the lower bound is the West Health/Gallup 11% unsafe-recommendation rate from real-world self-report.
[3]NEJM AI (Shekar, Pataranutaporn, Sarabu, Cecchi & Maes) / MIT Media Lab — People Overtrust AI-Generated Medical Advice despite Low Accuracy
Revisado por pares
Across 300 participants, people were unable to distinguish AI-generated from doctors' responses, rated high-accuracy AI responses as more valid/trustworthy/complete, gave low-accuracy AI responses ratings similar to doctors', and showed a high tendency to follow potentially harmful AI advice
Trecho
“"Participants were unable to effectively distinguish between AI-generated responses and doctors' responses [and] rated high-accuracy AI responses as significantly more valid, trustworthy, and complete than the other two types of responses. Low-accuracy AI responses tended to receive ratings similar to those given to doctors' responses [and participants showed] a high tendency to follow the potentially harmful medical advice contained in those responses."
”
Dados da fonte de
2025-05-13
Acessado
2026-05-11
Cálculo
NEJM AI / MIT Media Lab study on AI medical response evaluation. The overtrust finding is the mechanism explanation: users systematically overestimate AI reliability in the medical domain, meaning the problematic-response rates from accuracy studies translate into real-world harm more readily than they would if users could identify unreliable responses. This supports using the problematic-response rate (not just the unsafe- recommendation rate) as the relevant risk bound.
Fontes: inacção
Registro de evidências
Cada número abaixo é o que cada fonte relatou, com a citação literal em que nos baseamos e como chegamos ao nosso valor. Clique em qualquer link para verificar diretamente.
[1]West Health / Gallup — Millions of Americans Now Consult AI Before, After, and Sometimes Instead of Seeing a Doctor
Fonte de referência
46% of AI health users felt more confident asking providers questions afterward; 59% used AI to research before a doctor visit
Trecho
“"46% of AI health users felt more confident asking providers questions after using AI. 59% used AI to research before a doctor visit. 71% were motivated by wanting answers quickly; 71% wanted additional information."
”
Dados da fonte de
2026-01-01
Acessado
2026-05-11
Cálculo
West Health / Gallup panel, n=5,660, October–December 2025. The 46% who felt more confident asking providers questions is a concrete stated benefit of AI consultation — a benefit that non-users forgo. The 28% inaction-regret proxy is derived conservatively from the proportion of AI users who report concrete preparation or information benefits (46–59%), adjusted downward to reflect that non-users may have obtained similar information through other channels (search, calling a nurse line, reading drug inserts). No direct "do you regret not consulting AI for this decision?" survey was identified. This is the most data-sparse side of the entry.
[2]ABA Banking Journal (reporting on Wells Fargo and TD Bank surveys) — Bank Surveys Find Consumers Increasingly Turning to AI for Financial Advice
Fonte de referência
~90% of US adults who used AI for financial decisions and acted on the advice said results were 'profitable or worthwhile' (Wells Fargo 2026); 55% of adults use AI for financial management decisions (TD Bank 2026)
Trecho
“"19% of U.S. adults used AI for financial advice; 38% among Gen Z. Two-thirds of those who used AI acted on its suggestions. Approximately 90% of those who acted said the results were 'profitable or worthwhile.'"
”
Dados da fonte de
2026-04-01
Acessado
2026-05-11
Cálculo
Wells Fargo consumer survey and TD Bank consumer survey, both reported April 2026. Sample sizes and full methodology not disclosed. The 90% "worthwhile" self-report is the most positive available outcome figure for AI financial consultation but is highly susceptible to self-serving bias: people who acted on AI advice and lost money are less likely to report the action as worthwhile, but are also less likely to be included in a consumer satisfaction survey. Used here as directional corroboration that AI financial consultation produces positive self-reported outcomes at high rates, supporting the inaction-regret proxy. Not used as a primary figure due to methodological opacity.
Ressalvas
MEDIDAS POR PROXY EM TODO O CASO. Nenhuma pesquisa perguntou diretamente aos indivíduos "você se arrepende de ter consultado IA para esta decisão?" ou "você gostaria de ter consultado IA antes de decidir?" Ambos os lados são construídos inteiramente a partir de dados adjacentes: estudos de qualidade de resultado (para risco do lado da ação), benefícios relatados (para custo de oportunidade do lado da inação) e medidas comportamentais (taxas de não comparecimento a consultas médicas). O regret_delta de -0,06 é pequeno demais para classificar de forma confiável como inação dominante; é classificado como 'misto' porque a evidência não apoia claramente nenhum padrão. O risco do lado da ação varia drasticamente por domínio e por como a IA é usada: complementar uma consulta médica (baixo risco, alto benefício) vs. substituí-la para um sintoma sério (alto risco). Os estudos acadêmicos de precisão (taxas de resposta problemática 21-49%) refletem o desempenho da IA em perguntas médicas especificamente — o domínio de maior aposta. Decisões financeiras e de vida geral provavelmente têm taxas de resposta problemática mais baixas, mas menos benchmarks de precisão publicados. O estudo NEJM AI sobre confiança excessiva é uma ressalva particularmente importante: o mecanismo que converte respostas problemáticas em dano é a incapacidade dos usuários de detectá-las, e o enquadramento confiante com poucas ressalvas (documentado no estudo UCLA/BMJ Open) sistematicamente oculta essa incerteza. "Usar IA" não é uma única ação: ler um resumo de IA para se preparar para uma consulta médica é categoricamente diferente de agir sobre um diagnóstico de IA em vez de ver um médico. Esta entrada agrega esses usos heterogêneos, o que limita sua precisão. Os dados de auto-relato do Wells Fargo e do TD Bank são opacos sobre a metodologia e devem ser tratados apenas como direcionais. As capacidades da IA estão melhorando rapidamente; benchmarks de precisão e confiabilidade de 2025-2026 provavelmente serão diferentes dos aplicáveis 12-24 meses depois. Esta entrada tem uma vida útil mais curta do que a maioria no conjunto de dados.
