Posted on 05/12/2026 9:17:55 AM PDT by spintreebob
The civic tech nonprofit Code for America is partnering with artificial intelligence company Anthropic to develop tools aimed at helping caseworkers enhance public benefits administration across the nation.
The organizations are working together to develop an AI-enabled solution to improve the accuracy and timeliness of benefits service delivery under the Supplemental Nutrition Assistance Program, Jana Rhyu, vice president of product at Code for America, announced Friday at a summit hosted by the organization in Chicago last week.
The SNAP Policy Navigator tool is built on federal regulations, state manual selections, official policy directives and other documents to help caseworkers “quickly and accurately get an answer to [a] very specific policy question” when they are working with clients, said Michael Lai, who leads state and local government AI at Anthropic.
(Excerpt) Read more at nextgov.com ...
Dear FRiends,
We need your continuing support to keep FR funded. Your donations are our sole source of funding. No sugar daddies, no advertisers, no paid memberships, no commercial sales, no gimmicks, no tax subsidies. No spam, no pop-ups, no ad trackers.
If you enjoy using FR and agree it's a worthwhile endeavor, please consider making a contribution today:
Click here: to donate by Credit Card
Or here: to donate by PayPal
Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794
Thank you very much and God bless you,
Jim
"accuracy and timeliness" Deloitte IES is FRAUD. It has the DEAD eligible for benefits. DEAD is fraud. It has DUPLICATES. Sue Smith, Susan Smith, Susan Jones-Smith have the same DOB, SSN, address and are the same person. But she has 3 ID, 3 benefit plans in Deloitte IES. This is FRAUD. Many Addresses are uncleansed, unstandardized. Most but not all can be cleansed and standardized. Some are easy & cheap to fix. Some not so easy or cheap but can be fixed. DEI is inaccurate. A 3 tier table could store the data of Asian, E Asian, Japanese, Chinese, Korean. But 1/3 are UNKNOWN and hence any statistics are meaningless. May it is best to forget AI.
"Service Delivery" One vendor (eg Deloitte) handles Eligibility for Medicaid, CHPS, SNAP, WIC, TANF, Sec8, LIHEAP, etc. The CASE WORKERS work with the recipients and Deloitte data. Case workers do not (often) work with Providers or service delivery.
GAINWELL Technologies is the big dog in service delivery. Gainwell is not perfect. But the biggest WASTE FRAUD ABUSE in GAINWELL is working with garbage data it receives from DELOITTE IES.
Acentra, Optum, etc often processs the Provider data. Consulting has always driven the motto: If centralized, decentralize,. If decentralized, centralize. Always do the opposite of the current way. That generates the most consulting dollars.
GA Medicaid (and others) is pushed by BIDEN HHS to further decentralize GAINWELL into many more vendors.
The problem is not that systems are centralized or decentralized. The problem is the Deloitte iES is intentaionally FRAUD. Its EMPI Enterprise Master Person Index of IES does not meet any defintiion of Master Person Index and is therefor FRAUD.
AI, whther ANTHROPIC or other is GI/GO Garbage In Garbage Out. If CODE for America staff are at all competent they will quickly '1) Reaffirm the recommendation that Deloitte be charged and prosecuted for FRAUD by DOJ with monetary dmages. 2) The competent staff fix eligibility systems to fix the known errors. There are other errors in DELOITTE IES less major and/or too complex to cover here. They are often the cause of the much downstream WASTE FRAUD ABUSE.
(That Deloitte is the #1 abuser of H1b and that many IES H1b are not knowledgeable in US federal and state law and custom and many H1b are not competent in IT may be a different issue to discuss.)
Even Microsoft said never trust AI with anything important ,LOL
Detecting SNAP fraud seems to be a simple issue to solve if the people handing out SNAP ran the applicants name thru a database like USDA just did to detect SNAP recipients who were driving luxury cars.
Current AI data is good to develop “Leads, possibilities”.
Current AI has many problems. Some of them could be significantly reduced. But the priority seems to be elsewhere.
A narrow AI, a verry narrow AI, is more trustworthy than AI intended to impress the corporate and government buyers with other people’s money.
It takes less than a minute to write the SQL that will expose duplicates in a table.
The big problem with AI will be the security vulnerabilities.
The Deloitte IES database in 25 states with Deloitte in other states with sister products is not close to clean.
Running Deloitte data thru one of the many private sector databases would be much better as they have an incentive to be cleaner. State Farm, Allstate...none of the 35+ private sector database shops I have been in would accept the quality level of government work.)
Running the Deloitte data thru the SS DB would be a handy cross reference...if you think the SS DB ... at least for some. Likewise through the Unemployment Database. Unfortunately in in states like CA Deloitte does the unemployment database also.
There are a bunch of “African-American” government public union worker types who will squash it by claiming “AI be rayciss”[sic] ...
I have to admit that I spend an inordinate amount of time writing AI prompts and following AI generated rat holes where inquiries tend to lead you in completely unintended directions. My prediction is that Claude will lead to more wasted time and less productivity for the fraudsters at Deloitte IES. So, this is likely a good thing for US taxpayers who are currently being ripped of at an unprecedented rate even after Elon and DOGE tried to make a dent in this.
It takes less than a minute to write the SQL that will expose (SOME) duplicates in a table. Many different people have the same name. Many SSN are missing. SSN is not required. Some (eg Newborns) don’t have an SSN. Some SSN are DUPS (and many, but not all, are frauds). MANY ADDRESS are so bad, some SO BAD it is a major project to clean them, Often other deta in the data base that “might help narrow it down” doesn’t narrow it down.
Maybe you work for State Farm or Allstate and not the government... or maybe you don’t realize how bad the government data can be.
It takes less than a minute to write the SQL that will expose (SOME) duplicates in a table. Many different people have the same name. Many SSN are missing. SSN is not required. Some (eg Newborns) don’t have an SSN. Some SSN are DUPS (and many, but not all, are frauds). MANY ADDRESS are so bad, some SO BAD it is a major project to clean them, Often other deta in the data base that “might help narrow it down” doesn’t narrow it down.
Maybe you work for State Farm or Allstate and not the government... or maybe you don’t realize how bad the government data can be.
H1b at Deliotte (google their sites) are not US government union... and probably not union scale.
I have over eight years of hands-on experience dealing with government databases.
I would pay President Trump to give me the keys to one of those databases that are being exposed as messy.
from https://codeforamerica.org/news/anthropic-partnership/
“AI Tools for Caseworkers: New Claude integrations will enable caseworkers to process cases efficiently and accurately while navigating a complex and rapidly changing benefits landscape”
1) Under Obama welfare became larger, more complex and rapidly changing. Not 2025-26. Deloitte Integrated ELIGIBILITY System IES was designed and developed with full knowledge of the Obama changes and INTENTIONALLY designed and developed it to be a FRAUD.
2) GIGO Garbage In Garbage Out. The caseworkers receive garbage data.
3) I feel it necessary to say Gainwell Technologies and other IT vendors are not perfect. But their #1 WASTE FRAUD ABUSE in processing garbage data they receive from Deloitte.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.