Would you allow AI to access your bank account?
If an AI company wants to build a data center near your home, would you want them to disclose their water and electricity usage?
The closer AI tries to get into your life, the more it needs your trust. The question is: what does it offer in return?
Anthropic has launched a test version of Claude’s “Personal Finance” feature that can read your bank transaction records in a read-only mode and help you see where your money is going.
OpenAI has committed to no longer requiring local officials to sign NDAs (non-disclosure agreements) when building data centers and will publicly share water, electricity, and noise usage information.
Today’s two major AI news stories look at how AI companies are trying to gain public trust from two angles: personal data and community relations.
1. Claude Introduces Personal Finance Feature: Is It Safe for AI to Access Your Bank Account?
What does this feature do?
TechCrunch reported that on November 5 (US time), Anthropic launched the test version of Claude’s “Personal Finance”.
Users can connect their bank accounts and credit cards to Claude via the financial data aggregator Plaid, then directly ask questions about their finances.
According to TechCrunch and Plaid’s official blog, Claude can:
- Organize expenses by category.
- Identify potential issues such as duplicate charges, unexpected fees, or increasing bills.
- Explain terms like APR and calculate how much interest you’ve paid over the past year.
- Help plan the order in which to pay off debt.
Can Claude access or move my money?
Anthropic states that access is read-only; Claude cannot transfer funds, manage, or invest your money.
TechCrunch reports that Claude can see data including bank and account names, account types, last four digits of account numbers, balances, credit limits, and transaction history up to two years.
Anthropic also says conversations in this personal finance area will not be used to train the AI models. This feature has no ads, no sponsored recommendations, and Anthropic does not promote its own credit cards or loans through it.
Is it available in Taiwan?
Not at this time.
TechCrunch notes that the test version is only open to paid Pro, Max, and Ultra subscribers in the United States.
Plaid’s official blog also only mentions US consumers can link accounts.
Does “read-only” mean it’s risk-free?
Not necessarily.
Read-only means AI cannot move your funds, but sensitive information like transaction records and balances still pass through Plaid and Anthropic’s systems.
TechCrunch does not specify data retention periods or security certifications, and Plaid’s blog only provides a link to request data deletion.
Additionally, AI’s detection of “duplicate charges” or “abnormal fees” could be wrong. Any disputes or refunds still require you to verify with your original bank records.
Do other AI companies offer similar features?
TechCrunch mentions OpenAI introduced a similar ChatGPT personal finance tool in May 2026.
Plaid says over 800 AI companies already use its services. AI reading personal financial data is becoming common.
What can ordinary users do for now?
Even without linking bank accounts, you can safely have AI help organize your finances by:
First blacking out account numbers, names, and other personal info on statements or receipts, then letting AI categorize and sort them. See SasaDaily’s article “3 Steps to Organize Receipts Using ChatGPT” for guidance.
Sources:
- Plaid Official Blog: Plaid Powers Financial Insights In Claude
- TechCrunch: Claude can now analyze your spending with its new Personal Finance tool
2. OpenAI Promises No NDAs When Building Data Centers: Can This Ease Residents' Concerns?
What did OpenAI announce?
The Neuron reported that on November 5 (US time), two days after the US midterm elections, OpenAI released a "National Community Compact" and a policy blueprint for computational campuses aimed at policymakers.
The Compact is OpenAI’s commitment regarding its future campuses; the blueprint is a set of suggested rules for governments and the broader industry.
What does “no NDAs” mean?
In the past, tech companies often required local government officials to sign NDAs when negotiating data center projects, making it difficult for residents to learn the details.
The Neuron quotes the OpenAI Compact stating: OpenAI will not require public officials to sign NDAs and will not make NDAs a prerequisite for discussions or negotiations.
Exceptions apply to legitimate cybersecurity info, trade secrets, or competitively sensitive data.
OpenAI also said it canceled the NDA with Effingham County, Georgia, back in July.
What else did they promise?
The Neuron summarizes OpenAI’s commitments as:
- Publicly sharing water usage, electricity consumption, construction impacts, and economic contributions of each campus.
- Conducting independent noise studies before operation and annually thereafter.
- Separating water reports into “initial fill” and “ongoing usage”; economic reports into “estimated” and “actual.”
- Covering the cost of electricity and any new grid infrastructure for campuses, preventing cost shifts to other users.
Does this apply to all OpenAI data centers?
No.
The Neuron points out the Compact applies only to future campuses owned and operated directly by OpenAI. Facilities run by partners are not automatically covered.
The policy blueprint is a recommendation for governments and has no legal force.
Why announce this now?
SasaDaily reported yesterday that at least 14 towns in Ohio voted in the midterms to limit or ban data centers.
The Neuron also noted various communities backed restrictions on data center development in the election.
OpenAI’s timing — two days post-election — strongly suggests this is a response to rising opposition. This is an inference based on the timeline; OpenAI did not explicitly say so.
Are these promises effective?
Transparency and covering electric bills address key resident concerns: lack of openness and fears of cost shifts.
But whether these promises are fulfilled depends on actual data published for each campus going forward.
A company announcement alone does not mean issues are resolved. A more practical view is that residents and local governments now have a checklist to monitor performance.
Sources:
3. Why Are These Two Stories Worth Reading Together?
One story is about AI getting into your wallet; the other is about AI entering your community.
What both companies are really doing is:
Exchanging "clear rules" for ordinary people’s trust.
- Anthropic emphasizes read-only access, no training, and no advertisements to encourage account linking.
- OpenAI stresses no NDAs, data transparency, and paying for electricity to make communities more accepting of data centers.
Setting clear rules is a good start, but trust ultimately depends on delivering on those promises.
4. What Should Ordinary People Watch Out For Today?
If you’re considering letting AI access your financial data:
First clarify four things: Can AI move your money? Will your data be used for training? How long is your data stored? How can you delete it?
Financial advice from AI should only be a reference; any disputes or repayment decisions require checking your original bank records.
If your area is planning a data center or large facility:
Ask your local government: Did they sign NDAs? Are water, electricity, and noise data publicly available?
A company’s promises only matter if the public can verify them.
The Most Important Sentence for Today
AI companies want your trust—they must lay out the rules upfront; your job is to see whether they actually follow through.