AI in Biopharma Manufacturing: Are We There Yet?

Convergence: Where Is This Going — and Are We There Yet?

Sun, 06 Sep 2026 01:39:31 +0000

A site head wants one slide by Friday: four AI systems, three columns, do now, wait for, do anyway. Sam and Sarah close the season by reading the two documents that say where the rulebook is heading: the joint EMA–FDA Guiding Principles of Good AI Practice (January 2026), ten principles on two pages that put context of use, model risk, fit-for-use data, lifecycle and the human–AI team under both logos for the first time; and ICH's October 2025 reflection paper on advanced manufacturing, which names process modelling and continuous process verification as topics, lists the gaps in the 2011 Points to Consider in its own words, and proposes a process-models guideline as the first step, with no topic adopted yet. Then the calendar as the landscape verified it (CDER's promised manufacturing guidance, the Annex 22/11/Chapter 4 finals, the FDA draft, the AI Act's deferred dates), the EU machinery behind it (EMANS 2028, the NDSG and QIG workplans), the sandboxes that exist and the GMP sandbox that doesn't, the six open questions as the season found them, a scorecard for the four recurring systems, and Sam's verdict: nearly there on vocabulary, not on instruments, and the plant floor decides.

Transcript

Chapters

00:00 Intro
01:01 Cold open
02:11 Convergence: the question and the map
05:03 Joint principles: the shared vocabulary
09:18 ICH: the likely destination
18:23 The calendar: what is actually dated
22:05 Europe's machinery: EMANS, NDSG, QIG
25:11 Sandboxes: the Airlock and Article 57
28:43 The open questions, in one list
31:21 Scorecard: four systems, three columns
36:00 Are we there yet? Sam's verdict
37:48 So what for biomanufacturing
40:00 Recap: the season in three
41:11 The reading list

Documents discussed

Reading list, in order

Fact check

88 factual claims in this episode were checked against the source documents by an independent AI pass: 68 are supported by the primary text, 9 are confirmed by official web sources, 3 rest on secondary sources or are hedged as such (listed in the report), 8 corrected before release, 0 open. Full report.

Researched, scripted and voiced by AI systems under Jack Prior's direction. Sam and Sarah are AI characters; nothing in the episode is Jack speaking, and none of it is a statement of his views or his employer's. Generative AI can be confidently wrong — check the sources. Corrections: jack@jackprior.ai.

Change: What Happens When the Model — or the Process — Changes?

Sun, 06 Sep 2026 01:24:23 +0000

A vision vendor pushes a new model version at two in the morning and a language-model provider deprecates the version a site validated. Sam and Sarah split 'change' into four kinds (the environment drifts, the owner retrains, the model learns, the vendor pushes) and read the instruments inside out, anchored on ICH Q12: Q10 and the Annex 11 and Annex 22 drafts inside the quality system; Q12's reporting categories, Established Conditions and the post-approval change management protocol in the filing; FDA's 2025 draft tying them together with a lifecycle maintenance plan; and the device-side predetermined change control plan as the template nobody has married to the PACMP. Then the two questions FDA asked in 2023 and ICH repeated in 2025, a tier-by-tier rule for which instrument applies, the vendor update no document covers, and the four recurring systems taken through all four kinds of change, ending with what an AI PACMP for the soft sensor would actually say.

Transcript

Chapters

00:00 Intro
01:01 Cold open
02:05 Four kinds of change, why now
04:26 Four kinds of change, not one
07:01 Inside the PQS: Q10, Annex 11, Annex 22
10:50 Q12: Established Conditions and reporting
13:25 PACMP: a change agreed in advance
15:51 FDA: the lifecycle maintenance plan
19:01 PCCP: the device template
21:51 The AI PACMP: marrying the two
24:03 Two open questions and the tier rule
27:32 The vendor update, and two asymmetries
33:12 Four systems, four changes
37:07 So what for biomanufacturing
38:57 Recap and next time

Documents discussed

Fact check

68 factual claims in this episode were checked against the source documents by an independent AI pass: 63 are supported by the primary text, 2 are confirmed by official web sources, 1 rest on secondary sources or are hedged as such (listed in the report), 2 corrected before release, 0 open. Full report.

Humans: Who Is Accountable When the Model Is Wrong?

Sun, 06 Sep 2026 00:58:51 +0000

Every document that lets AI into GMP lets it in on one condition: a human in the loop. Sam and Sarah read what that phrase actually means across six documents, anchored on EMA's 2024 reflection paper: BioPhorum's three oversight settings and the grade each one earns, the EU AI Act's Article 14 list of what an overseer must be able to do, draft Annex 22's operator monitored like any other manual process, FDA's and the GMLP principles' human-AI team, and the paradox the documents state themselves, that oversight is both a control and a risk. Then Sam's question, parked since episode four: which human, with which competence, evaluating which inputs, measured how. The four recurring systems get the three questions, the agentic assistant gets the night-shift test, and Sam's position is offered for the listener to test: encode every check a computer can run, name the human judgment by competence, and if that competence is not on shift, the control is fiction and the grade must not get credit for it.

Transcript

Chapters

00:00 Intro
01:01 Cold open
02:19 Which human, and why now
04:58 Three meanings of human oversight
07:17 AI Act Article 14: what an overseer must do
09:57 Annex 22: the operator as a manual process
13:13 FDA and GMLP: evaluate the human-AI team
15:51 EMA: a risk plan for every model
18:59 The oversight paradox
21:16 Which human, which competence, which inputs
24:23 Accountability does not move
25:56 Sam's position: encode, name, escalate
27:54 Four systems, three questions
33:09 So what for biomanufacturing
35:22 Recap and next time

Documents discussed

Fact check

47 factual claims in this episode were checked against the source documents by an independent AI pass: 43 are supported by the primary text, 2 are confirmed by official web sources, 0 rest on secondary sources or are hedged as such (listed in the report), 2 corrected before release, 0 open. Full report.

Data: Are My Data Fit for Use?

Sun, 06 Sep 2026 00:44:51 +0000

Every framework this season assumes the data under the model are trustworthy. Sam and Sarah read what the rulebook actually requires: ALCOA and its pluses from FDA's 2018 Data Integrity Q&A, PIC/S PI 041 and the Annex 11 revision; FDA's 2025 'fit for use', which adds relevance to reliability; and the data-governance clauses of the AI Act, EMA's reflection paper and draft Annex 22. The Annex 11 revision turns the training-data pipeline into a GMP computerised system, with validated transfers, unique accounts, audit trails and backup. Then the gap: nobody says how to get from a historian, a batch record and a LIMS to a fit-for-use dataset. The soft sensor's training set is interrogated attribute by attribute, the batch model's golden batches become exclusion decisions, the camera's labels and vials get their clauses, and the agentic assistant's data turn out to be documents, versions and all.

Transcript

Chapters

00:00 Intro
01:01 Cold open
01:57 Are my data fit for use, and why now
05:11 Three definitions of trustworthy data
07:27 Annex 11: data handling becomes GMP
11:40 FDA's Data Integrity Q&A in one breath
16:21 Relevance and provenance: the AI texts
20:43 The pipeline gap nobody writes down
23:57 The soft sensor's training set, attribute by attribute
27:14 Batch model, camera and the agent's documents
31:04 So what for biomanufacturing
33:34 Recap and next time

Documents discussed

Fact check

69 factual claims in this episode were checked against the source documents by an independent AI pass: 65 are supported by the primary text, 4 are confirmed by official web sources, 0 rest on secondary sources or are hedged as such (listed in the report), 0 open. Full report.

Evidence: What Do I Have to Show?

Sun, 06 Sep 2026 00:29:35 +0000

Once a model is allowed through the gate, what do you owe on paper? Sam and Sarah read FDA's 2025 draft for its credibility assessment plan and report, sized to model risk, and then read draft EU GMP Annex 22 sections three to nine as the most concrete evidence recipe in the corpus: an intended use owned by a process expert, acceptance criteria fixed before testing and no lower than the process being replaced, a designed test set with verified labels, test-data independence enforced by access control and staff separation, and explainability and confidence recorded as evidence. FDA's Computer Software Assurance guidance, the ten Good Machine Learning Practice principles and the 2023 credibility vocabulary explain how validation became assurance. The fill-volume camera's evidence package is built clause by clause, then the soft sensor, the batch model and the agentic assistant go through the same recipe.

Transcript

Chapters

00:00 Intro
01:01 Cold open
02:08 What do I have to show, and why now
05:00 FDA: plan, execute, report, commensurate
07:28 Annex 22: intended use owned by a process expert
10:55 Annex 22: test data and test execution
13:31 Independence is a people-and-access control
16:20 Explainability and confidence become evidence
18:10 Five FDA asks MSAT would not expect
21:11 From validation to assurance: CSA and GMLP
26:31 The fill-volume camera's evidence package
30:56 Soft sensor, batch model and the agent
35:33 So what for biomanufacturing
37:59 Recap and next time

Documents discussed

Fact check

58 factual claims in this episode were checked against the source documents by an independent AI pass: 52 are supported by the primary text, 4 are confirmed by official web sources, 1 rest on secondary sources or are hedged as such (listed in the report), 1 corrected before release, 0 open. Full report.

Model Type: Can It Learn After Deployment — and Can It Be an LLM?

Sun, 06 Sep 2026 00:04:07 +0000

For twenty years the model you picked was an engineering choice. Draft EU GMP Annex 22 makes it a compliance decision: static and deterministic machine learning only in critical GMP applications, with dynamic models, probabilistic models and generative AI told they should not be used there at all. Sam and Sarah read the six-page draft in full, set it against FDA's 2025 draft, which tells sponsors to anticipate self-evolving models and never mentions language models, and against the device world's predetermined change control plans. EMA's reflection paper and NIST's generative AI profile supply the vocabulary for the risks. The four recurring systems go through the gate, and the episode ends with how a site keeps a retraining loop outside the GMP boundary.

Transcript

Chapters

00:00 Intro
01:01 Cold open
02:05 Why model type is now a compliance question
04:44 Annex 22: provenance, maturity, narrow scope
06:43 Static and deterministic: the gate by name
09:38 Generative AI: non-critical only, human in the loop
12:31 FDA: anticipate the self-evolving model
15:29 PCCP: the device world's answer
17:56 FDA's silence and the operational-efficiency carve-out
19:58 EMA and NIST name the generative risks
24:02 One chapter of a larger rewrite
26:05 Four systems through the gate
31:57 So what for biomanufacturing
35:05 Recap and next time

Documents discussed

Fact check

64 factual claims in this episode were checked against the source documents by an independent AI pass: 58 are supported by the primary text, 4 are confirmed by official web sources, 2 rest on secondary sources or are hedged as such (listed in the report), 0 open. Full report.

Grading: How Much Does This Model Matter?

Sat, 05 Sep 2026 20:32:40 +0000

Seven documents grade models seven ways, and they grade different things. Sam and Sarah start with the 2011 ICH Points to Consider, which graded models low, medium and high by their role in assuring quality and made 'sole indicator' the line, then set beside it FDA's and M15's model risk, Annex 22's critical-or-not test and model-type gate, the AI Act's product tiers, and BioPhorum's redefinition of model influence as maturity. Four tiers emerge for the season. All four recurring systems are placed and moved, and the agentic assistant's tier turns out to be an argument about criticality that has to be won first.

Transcript

Chapters

00:00 Intro
01:01 Cold open
01:59 Orientation: seven gradings
04:09 ICH 2011: low, medium, high impact
07:39 Q9: formality scales with risk
08:33 Impact is not impact
09:57 Annex 22 grades the application
13:15 The AI Act in one takeaway
14:39 BioPhorum redefines model influence
18:49 Four tiers for the season
21:52 Four systems, placed and moved
24:27 The agent's tier
26:35 So what for biomanufacturing
29:13 Recap and next episode

Documents discussed

Fact check

69 factual claims in this episode were checked against the source documents by an independent AI pass: 62 are supported by the primary text, 6 are confirmed by official web sources, 1 rest on secondary sources or are hedged as such (listed in the report), 0 open. Full report.

Context of Use: What Decision Is the Model Making?

Sat, 05 Sep 2026 19:59:46 +0000

The one idea in nearly every document: context of use. Sam and Sarah read FDA's January 2025 draft closely, from question of interest to role and scope to model risk as influence times consequence, trace the vocabulary from ASME's device standard through FDA's 2023 device guidance to ICH M15, and set EMA's two axes beside it. FDA's fill-volume camera is medium risk only because a sample test stays in place. All four recurring systems go through steps one to three, and the agentic assistant turns out to be the one whose scope sentence is hardest to write honestly.

Transcript

Chapters

00:00 Intro
01:01 Cold open
01:39 Orientation: the one shared idea
04:06 The decision, not the model
07:38 Lineage: ASME to FDA's device guidance
08:56 ICH M15's four terms
10:58 The fill-volume example
13:29 Seven steps and early engagement
16:19 EMA's two axes
19:29 Four systems through steps one to three
22:03 The agent's scope sentence
24:49 So what for biomanufacturing
27:23 Recap and next episode

Documents discussed

Fact check

74 factual claims in this episode were checked against the source documents by an independent AI pass: 70 are supported by the primary text, 4 are confirmed by official web sources, 0 rest on secondary sources or are hedged as such (listed in the report), 0 open. Full report.

Five Trends: How the Expectations Are Evolving

Sat, 05 Sep 2026 19:47:23 +0000

Before reading any single document, you need the map. Sam and Sarah sort the AI-in-manufacturing rulebook into its four layers, replace "binding or not" with maturity and who-will-ask, and trace five trends across the documents: from models to AI, from model-agnostic to Annex 22's line on model type, from validation to lifecycle, from regional to the joint EMA-FDA principles and ICH's reflection paper, and from silence on generative AI to a gap everyone can see. The recurring cast is introduced, and the agentic assistant that drafts deviation investigations is walked down the maturity axis to show how little the documents yet say about it, and how a site reasons about it anyway.

Transcript

Chapters

00:00 Intro
01:01 Cold open
01:57 Orientation: whose expectation, how settled
04:56 Four layers, four vocabularies
06:44 Non-binding is not a strategy
08:53 Trend one: from models to AI
12:28 Trend two: Annex 22 breaks ranks
15:13 Trend three: validation to lifecycle
18:31 Trend four: the joint principles
20:44 ICH's reflection paper
22:50 Trend five: silence on generative AI
24:31 The four recurring systems
26:33 The agent on the maturity axis
32:09 So what for biomanufacturing
34:57 Recap and next episode

Documents discussed

Fact check

148 factual claims in this episode were checked against the source documents by an independent AI pass: 121 are supported by the primary text, 20 are confirmed by official web sources, 6 rest on secondary sources or are hedged as such (listed in the report), 1 corrected before release, 0 open. Full report.

Are We There Yet? (Trailer)

Sat, 05 Sep 2026 19:30:00 +0000

Jack Prior introduces the show: a bioprocess engineer who started out in the late eighties at MIT looking at how to apply AI to bioprocesses, on why 2026 is the year that time has arrived — and the question the podcast asks of our data readiness, of machine learning and agentic AI, and of the regulatory frameworks: are we there yet? Season 1 works through the forty-odd guidances and white papers that make up the regulatory landscape, in a narrative format produced with AI and fact-checked against the source documents. Everything at https://jackprior.ai/podcast

Transcript