Federal mandate M-23-07: Leverage AI for efficient compliance
From ordering Thai takeout to refinancing your home online, digital transformation is revolutionizing just about every facet of society. In the age of IT modernization and new ways of working, organizations around the world are leaving outdated paper-based processes behind in favor of digital methods.
Anne Siess
IT professional
Slowly but surely, these changes are coming to government agencies. For federal agencies, a key driver of this transformation is the National Archives and Records Administration (NARA) M-23-07 directive, which has been updated from M-19-21. This directive requires all federal agencies to manage their permanent records electronically by June 30, 2024.
Let’s look at what NARA M-23-07 means for federal agencies and how the right tools can ease the transition — and make records management more effective than ever.
What is M-23-07?
M-23-07 is a federal mandate that came out in a December 23, 2022 memorandum. It directs all federal agencies to do two things:
- Ensure all federal records are created, maintained and managed in electronic formats, with appropriate metadata
- Develop plans to close agency-operated storage facilities for paper and other analog records, and transfer those records to federal records centers operated by NARA or commercial storage facilities
Hyland on AWS
The implementation deadline has been extended from December 31, 2022 to June 30, 2024.
Beginning July 1, 2024, “all agencies must transfer permanent records to NARA in electronic formats and with appropriate metadata in accordance with NARA regulations and transfer guidance,” according to the memo.
Why should federal agencies go paperless?
M-23-07 is part of the federal government’s goal to drive transformational change for a more modernized government. The heads of the Office of Management and Budget (OMB) and NARA noted these points in the memo for enacting this directive:
- Paper records management is costly to taxpayers and time-consuming.
- Security investments can be applied more effectively to managing electronic records.
- Digitizing processes creates a better experience for citizens, largely eliminating the use of cumbersome traditional methods like in-person interactions or mail to submit information such as applications or forms.
We asked agencies to create maturity models. We want to see agencies improving, addressing challenges, mitigating risks and getting to a higher level of maturity. We’re going to keep working on managing electronic records and trying to get to that environment where we’re fully electronic and realizing the vision of ‘zero click’ records management.
— Laurence Brewer, Chief Records Officer, NARA
Envisioning a post-mandate world
The M-23-07 mandate presents a significant cultural shift for the hundreds of federal agencies with long-standing paper-based processes for managing documents. Within just one agency, there are a multitude of content item types, processes and system integrations.
Agencies need to implement a strategy that enables them to not only effectively manage, organize and report on documents but also scale as their agency or content grows.
In the recent webinar, “Artificial intelligence and machine learning for records management,” Hyland Director of Federal Government Solutions Troy Doller shared his viewpoint for agencies looking to meet the mandate:
“One of the major benefits of this mandate is, it provides agencies an opportunity to build a lasting information management program that will benefit the agency moving forward,” Doller said.
The vision for a more modernized agency is there, but how can they attain it — quickly?
3 Ways AI Can Help You Meet the M-23-07 Transition to Electronic Records Mandate
To achieve compliance with the M-23-07 mandate, what’s the fastest path forward?
Automation solutions that leverage artificial intelligence (AI) and machine learning (ML) can improve current processes quickly and facilitate compliance. Hyland’s Alfresco Digital Business Platform (DBP) integrates with several Amazon Web Services (AWS) AI capabilities that can improve records management.
Read more to learn about three high-powered AWS tools.
1. Amazon Textract: AI can identify, capture and sort content
Paper-based agencies have an immense volume of paper records. These have been created daily, for years, and the volume of data has only grown.
Agencies making the switch to M-23-07 compliance need efficiency in their processes to serve their staff and stakeholders going forward. This can be achieved by:
- Identifying all content coming into the agency
- Capturing the content from its source — not just reactively scanning paper documents
- Transforming the captured data into usable, searchable electronic records
Agencies need a process in place to facilitate the transformation of the record to electronic means. One option for identifying content is leveraging an AI capability that performs document analysis. Amazon’s Textract provides this analysis by automatically extracting text and data from a variety of documents, including:
- PDFs
- Word documents
- Documents with handwriting
- Forms of all varieties
- Specialized formats (e.g., invoices, expenses and identity documents)
- Tables of all varieties
The metadata that comes out of Amazon Textract can be put in a searchable database that is made available within a content solution.
Derek Doerr, an AWS principal solutions architect and co-host of the “AI and ML for Records Management” webinar, said that without this type of intelligent automation, agencies will find it difficult to process records or make them usable and accessible.
2. Amazon Comprehend: AI capabilities can help users see a better future
User adoption can be a challenge for any digital transformation initiative.
Change is hard, and getting a government employee who has diligently followed the same processes day-in and day-out to use a new solution can be tough.
“For many government agencies, their biggest barrier will be getting employees to adopt the solution, get trained in the solution and be able to adapt when the solution rolls out a new update,” Doller said.
Often, seeing is the trick to believing.
When employees can visualize a digitized version of their current manual processes, they’re more likely to take part in bringing it to life. For example, a solution like Amazon’s Comprehend, which uses natural language processing (NLP) to digest content and extract insights, will eliminate time-consuming tasks for employees. Amazon Comprehend can:
- Surface the information workers need
- Infer relationships in unstructured data
- Extract data insights
- Make that data searchable
“Amazon Comprehend can look at a chunk of text and process it in different ways,” Doerr said. “For example, it can go in pretrained models to extract known entities. So it could pick out organization name, a location, dates and times, a person’s proper name, quantity and so forth. It can also go through and identify whether a piece of data looks like personally identifiable information and then process a redaction of it.”
3. Amazon Rekognition: AI can control information sprawl
Twenty years ago, there was much less content being created.
Today, information is created at a much faster pace and stored in a variety of areas. This not only compounds the problem of managing records in a variety of systems but also makes properly governing all that data manually an overwhelming, if not impossible, task.
Using AI in an automated process can be an effective solution. Amazon Rekognition, a service that can identify objects, people, text, scenes and activities in images and videos using its own intelligence, can help agencies keep up with information sprawl.
Amazon Rekognition adds this analysis to applications to help organizations catalog assets, automate media workflows and extract meaning.
“In a case where an agency or department has digital media beyond straight-up documents — such as audio or video recordings — those can all be processed and tasked to look for keywords, then extract them out and make them searchable and accessible,” said Doerr.
For example, Amazon Rekognition could “see” a photo that contains, say, a street sign and:
- Identify the text
- Extract the text in its entirety
- Classify the text
- Feed the text into Amazon Comprehend to give the text fuller context
This AI-enabled automation can take not just structured data but also unstructured data and intelligently organize it for future, actionable use.
Launch Your M-23-07 and AI Strategy with Hyland’s Alfresco Platform
MAI and ML are key enablers for driving operational efficiencies for records managements as well as in other agency processes that involve content management. Hyland’s Alfresco DBP has integrations with all the Amazon solutions discussed above — Amazon Textract, Amazon Comprehend and Amazon Rekognition — to transform and reduce the manual tasks that employees do every day.
To learn more insights from Troy Doller and Derek Doerr, listen to the recorded webinar, “Modernizing records management with intelligent automation.”
Hyland on AWS
Hyland is listed in the AWS Marketplace. Learn more about the benefits of purchasing there, including the ability to:
- Streamline procurement
- Implement controls and automate provisioning
- Manage software budgets with cost transparency