Latest / Deep Tech Germany – AI, Robotics & Frontier Innovation by Startuprad.io™ / Startup News Germany, Austria, Switzerland July 2026: The Structural Rotation

Transcript

Startup News Germany, Austria, Switzerland July 2026: The Structural Rotation · Deep Tech Germany – AI, Robotics & Frontier Innovation by Startuprad.io™

0:00
One point eight billion dollars. That is what Munich-headquartered Helsing
0:04
raised in a single Series E round in July.
0:08
The round valued the defence AI company at eighteen billion dollars.
0:13
In the same month, Quantum Systems announced a one point two billion
0:19
dollar Series D at an approximately eight billion dollar post-money valuation.
0:24
Proxima Fusion raised four hundred and eleven million euros at a valuation
0:29
of two point four billion euros.
0:32
Uber announced a voluntary takeover offer for Delivery Hero.
0:36
Eli Lilly agreed to acquire AtaiBeckley.
0:39
And SAP completed its acquisition of Freiburg-based Prior Labs after roughly
0:44
eighteen months of company-building.
0:46
These events do not all belong in the same accounting bucket.
0:51
Three are startup financing rounds. Two are proposed acquisitions that have
0:56
not yet closed. One is a completed acquisition with an undisclosed purchase
1:01
price. But together they show the same direction of travel: global
1:06
institutional and strategic capital is paying attention to German and
1:11
German-founded technology companies in defence, autonomy, fusion, biotech, and
1:15
applied AI.
1:16
My thesis is that July made a structural rotation visible.
1:21
Not a complete replacement of software by hardware, and not the end
1:26
of consumer technology. The shift is narrower and more defensible than that:
1:32
Germany's largest disclosed growth rounds are increasingly concentrated in
1:36
capital-intensive, strategically relevant technologies.
1:38
This is Startuprad.io. I am Joe Menninger, joining you from Frankfurt am
1:44
Main. You are listening to the July 2026 news cut for Germany,
1:50
Austria, and Switzerland. This is a solo episode.
1:55
The factual claims are sourced, and the interpretations and predictions are
2:00
mine.
2:01
Before we go further, let me check the predictions I put on
2:07
record in June.
2:09
Prediction one: STARK reaches a ten billion euro valuation within eighteen
2:14
months. Status: pending. STARK's five hundred million euro financing round, at
2:20
a reported valuation of three point two billion euros, was announced on
2:26
June twenty-third. It is an important part of the wider defence trend,
2:33
but it is June news and is not included in July's funding
2:39
total.
2:39
Prediction two: the KNDS IPO prices above fifteen billion euros and triggers
2:46
at least one more defence listing.
2:49
Status: missed on timing. On July first, the Franco-German defence group put
2:55
the planned Frankfurt and Paris listing on hold until market conditions
3:01
improve. The reported valuation range before the postponement was twelve to
3:06
fifteen billion euros, not above fifteen billion.
3:10
Prediction three: Isar Aerospace reaches orbit before December thirty-first,
3:15
2026. Status: pending. Isar Aerospace had not completed a successful orbital
3:20
flight by the end of July.
3:24
Deutsche-startups.de counted approximately three point seven billion euros
3:32
across eighteen disclosed double- and triple-digit-million investments in
3:41
July. Helsing, Quantum Systems, and Proxima Fusion were the dominant rounds.
3:52
The cleanest number for German defence-company financing in July is three
4:02
billion dollars: one point eight billion for Helsing and one point two
4:14
billion for Quantum Systems. STARK's five hundred million euro round belongs
4:25
to June. Lakestar's three hundred million dollar fund is a fund commitment,
4:36
not money raised by a startup.
4:42
And QuantumDiamonds' ninety-one million euro package combines only fifteen
4:51
million euros of equity with seventy-six million euros of non-dilutive public
5:01
funding.
5:02
Those distinctions matter. Company equity, secondary share sales, venture-fund
5:11
commitments, and public grants are all useful capital, but they are not
5:23
interchangeable.
5:24
PitchBook's first-half assessment points in the same broad direction: German
5:33
venture funding accelerated around deep tech and defence while deal count
5:44
declined. That is concentration, not a universal boom.
5:52
The global context is also concentrated.
5:58
Crunchbase reported a record five hundred and ten billion dollars in global
6:09
venture funding in the first half of 2026.
6:17
But OpenAI and Anthropic alone accounted for two hundred and seventeen billion
6:29
dollars, or forty-three percent of the total.
6:35
The global boom is real, but a small number of very large
6:47
companies drive a remarkable share of it.
6:54
Germany's version of that concentration is different.
7:01
The largest July rounds were not frontier-language-model rounds.
7:09
They were defence AI, autonomous systems, and fusion energy.
7:18
But there is an important counterweight.
7:24
The German Startup Association counted three thousand and fifty-three new
7:33
startups in the first half of 2026, the highest half-year total since
7:45
its series began in 2019. Software remained the largest category, with eight
7:57
hundred and forty-four new companies. Industry was the fastest-growing
8:05
category compared with the second half of 2025.
8:13
So the precise thesis is this: software still dominates company formation,
8:24
while hard-tech and strategic technologies increasingly dominate the largest
8:33
late-stage financing headlines.
8:36
Start with Helsing. On July thirteenth, the company announced a one point
8:43
eight billion dollar Series E at an eighteen billion dollar valuation.
8:50
Its investor group included Dragoneer, Lightspeed, ICONIQ, Growth Equity at
8:56
Goldman Sachs Alternatives, JPMorganChase, CPP Investments, General Catalyst,
9:01
Plural, and StepStone.
9:02
The Financial Times reported that Dragoneer led the round.
9:08
Growth Equity at Goldman Sachs Alternatives participated.
9:12
That distinction matters because participation and leadership are different
9:18
signals.
9:18
An eighteen billion dollar private valuation puts Helsing firmly in decacorn
9:25
territory. My read is that the investor mix shows defence AI and
9:32
software-defined military systems moving from specialist venture portfolios
9:37
into mainstream growth capital.
9:40
Quantum Systems announced its financing on July second.
9:45
The one point two billion dollar Series D valued the company at
9:52
approximately eight billion dollars post-money. Blackstone, Noteus, Airbus,
9:57
and Advent co-led the transaction, with additional participation from BOND,
10:03
Fidelity, Wellington, A.P. Moller Holding, Elephant Lake Ventures, and
10:09
existing shareholders.
10:10
Quantum Systems is not simply a hardware story.
10:15
It develops autonomous systems across air, land, and sea, connected by its
10:23
software and AI platform. The company said the financing would expand
10:29
production, strengthen its supply chain, support international delivery, and
10:35
fund further software and AI development.
10:39
There was also a significant shareholder-structure signal.
10:43
Co-CEO Florian Seibel told the Financial Times that investors uncomfortable
10:49
with a move into lethal systems had an opportunity to exit during
10:56
the transaction. That is evidence of a changing capital base: investors that
11:04
entered a dual-use company under one mandate can be replaced by investors
11:11
that explicitly accept defence as the thesis.
11:15
STARK reinforces the wider sequence, but it should be placed correctly on
11:23
the calendar. Its five hundred million euro round was announced on June
11:30
twenty-third at a reported valuation of three point two billion euros.
11:37
It is context for July, not July primary capital.
11:42
Finally, Lakestar closed its Resilience I fund on July seventeenth with three
11:50
hundred million dollars for European and NATO-allied dual-use and defence
11:56
technology. Lakestar described it as Europe's largest dedicated institutional
12:01
venture fund of its kind. Again, that is fund capital available for
12:09
future investments, not a startup financing round.
12:13
The conclusion survives the corrections. Two Munich-area defence companies
12:19
announced three billion dollars of financing within eleven days.
12:24
The round sizes and investor composition show that European defence technology
12:31
has become an institutional growth category.
12:35
Now to energy. On July seventh, Munich-headquartered Proxima Fusion announced
12:39
a four hundred and eleven million euro financing round at a two
12:44
point four billion euro valuation. XTX Ventures and East X Ventures led
12:49
the round. The company named Google and RWE as strategic investors, alongside
12:54
KfW Capital, SPRIND, Burda Principal Investments, the EIC Fund, and other new
12:59
and returning backers.
13:00
Proxima described the financing as the largest ever in European fusion.
13:04
It will use the money to build Alpha, its planned net-energy stellarator
13:09
demonstrator near Munich, and to expand its magnet, cable, engineering, and
13:14
manufacturing capabilities.
13:15
RWE's participation is especially concrete. Before investing, the energy
13:19
company had already signed an agreement with Proxima to work toward a
13:23
first stellarator fusion power plant at the former Gundremmingen nuclear site
13:28
in Bavaria.
13:29
Google's participation signals long-term interest in firm, carbon-free energy,
13:33
but it does not prove that Google expects commercial fusion on a
13:38
particular timetable. Fusion remains technically difficult and commercially
13:41
unproven. The accurate signal is that major technology and energy companies
13:45
now consider the option strategically important enough to finance.
13:49
That is the sovereignty argument in its strongest form: not that Europe
13:54
has solved fusion, but that it is funding the scientific and industrial
13:59
capacity required to find out whether it can.
14:03
Germany's exit market remains the weak point, and July produced a meaningful
14:07
pipeline. But two of the three headline transactions were agreements, not
14:11
completed exits.
14:12
First, Delivery Hero. On July sixteenth, Uber entered into a
14:16
business-combination agreement and announced a voluntary takeover offer of
14:19
forty-one euros and fifty cents per share.
14:22
Uber put the fully diluted equity value for one hundred percent of
14:26
Delivery Hero at fourteen point eight billion dollars, or thirteen point seven
14:31
billion dollars after adjusting for Uber's earlier stake purchases.
14:34
Delivery Hero's management and supervisory boards supported the offer at
14:38
announcement, and Prosus committed to tender its stake.
14:41
But the offer remains subject to its acceptance threshold and regulatory
14:45
approvals. Closing is expected in the second half of 2027.
14:49
Delivery Hero therefore had not been sold to Uber in July 2026.
14:53
If the transaction closes, it will be a major technology outcome from
14:58
Berlin. Until then, it is an announced takeover offer, not realised liquidity
15:02
for the full shareholder base.
15:04
Second, AtaiBeckley. On July fifteenth, the company signed a merger agreement
15:08
with Eli Lilly. The terms provide six dollars and seventy-five cents per
15:13
share in cash at closing, representing approximately two point eight billion
15:17
dollars, plus contingent-value rights worth up to another two dollars and
15:21
fifty cents per share, or approximately one billion dollars in aggregate, if
15:26
specified clinical and regulatory milestones are achieved.
15:28
The transaction still requires shareholder approval, regulatory clearances,
15:31
and other closing conditions. Lilly had agreed to acquire AtaiBeckley; it had
15:36
not yet paid three point eight billion dollars.
15:39
And the final value may be lower because the additional one billion
15:44
dollars is contingent.
15:45
Third, Prior Labs. This one did close.
15:47
SAP announced completion of the acquisition on July seventeenth after signing
15:51
the agreement in May. Prior Labs develops foundation models for tabular data
15:56
and will continue to operate as an independent entity within SAP's orbit.
16:01
The purchase price was not disclosed.
16:03
SAP separately committed to invest more than one billion euros over four
16:07
years to scale Prior Labs into a frontier AI lab.
16:11
That future investment commitment is not the acquisition price.
16:14
Prior Labs' founders said they had built the company over eighteen months
16:19
before signing the SAP agreement. That is a remarkably fast route from
16:24
founding to strategic acquisition, but there is no reliable basis for calling
16:28
it the fastest German AI exit in history.
16:31
The corrected exit thesis is therefore more cautious: July created two
16:35
potentially large future liquidity events and completed one strategically
16:39
important but undisclosed acquisition. Capital recycling depends on the
16:42
pending deals actually closing.
16:44
European fintech is not dead. It is selective, and regulatory execution
16:49
matters as much as valuation.
16:51
On July twenty-first, Augustus announced a one hundred and eighty million
16:56
dollar Series B at a one billion dollar valuation.
17:00
Tiger Global led the round. Augustus is building an API-first clearing bank
17:06
intended to give international fintechs and banks direct access to US-dollar
17:11
accounts and payment rails.
17:12
The regulatory caveat needs precision. In May, the US Office of the
17:18
Comptroller of the Currency granted Augustus National Bank preliminary
17:22
conditional approval. That is significant, but it is not final authorisation
17:27
to open. The OCC said final approval depends on Augustus meeting its
17:33
pre-opening requirements, including capital, operational, compliance, and
17:36
other regulatory conditions.
17:37
On July twenty-ninth, Berlin-based Pliant said it had surpassed one hundred
17:42
million dollars in annual recurring revenue.
17:45
That is a company-reported operating milestone, not a funding round.
17:50
Pliant said it served more than five thousand businesses and more than
17:55
thirty-five partners across Europe and the United States.
17:59
A post-recording update: on August fifth, Berlin-founded Moss announced a
18:04
thirty-five million euro Series C led by Portage at a one billion
18:09
euro valuation. The company said the financing would support AI tools that
18:15
finance teams can direct, inspect, and approve rather than fully autonomous
18:20
workflows. This belongs to August, not July.
18:23
And one further update is necessary for accuracy at publication.
18:27
In July, Reuters reported that Stripe and Advent International had offered
18:33
sixty dollars and fifty cents per share, or about fifty-three billion dollars,
18:38
for PayPal. On August twenty-eighth, Reuters reported that the consortium had
18:43
abandoned the pursuit. The July approach was real, but it is no
18:49
longer a pending transaction.
18:51
On July twenty-second, Germany's federal cabinet approved a new Startup and
19:00
Scaleup Strategy. The government said the strategy covers eight fields of
19:09
action, including financing, talent, faster company formation, public
19:15
procurement, internationalisation, and technology transfer. It also includes
19:22
security and defence startups systematically for the first time.
19:29
That is a real policy event.
19:34
But policy intent is not deployed venture capital.
19:41
The implementation, eligibility rules, procurement practice, and mobilisation
19:48
of private investors will determine the impact.
19:54
The pension debate needs the same discipline.
19:59
Germany is discussing a stronger capital-market component in retirement
20:07
provision. That does not mean one trillion euros is available for startups,
20:17
and it does not establish a direct five billion euro venture-capital pipeline.
20:27
I am not using those figures in this analysis.
20:34
Corporate capital is easier to measure.
20:39
On July twenty-third, Fresenius established Fresenius Ventures with an
20:47
intended investment volume of more than two hundred million euros over five
20:57
years. The unit will invest from early financing rounds through growth stage
21:06
in areas adjacent to biopharma, medtech, and care provision.
21:14
This is another form of the same institutional pull: established European
21:23
companies are building dedicated vehicles to access innovation earlier.
21:31
The month also contained evidence against an uncomplicated boom narrative.
21:37
KNDS postponed its planned IPO on July first, citing difficult conditions in
21:45
defence equity markets. Private investors were willing to finance Helsing and
21:52
Quantum Systems at large valuations, while public-market conditions stopped a
21:59
major defence listing. Private and public capital were not sending the same
22:07
signal.
22:07
Enpal confirmed that it would close its Hamburg sales location, affecting up
22:15
to eighty-five employees. This followed a separate customer-service
22:21
restructuring in Berlin earlier in the year.
22:25
At the same time, Enpal said first-half revenue exceeded six hundred million
22:33
euros and targeted more than one point three billion euros for the
22:41
full year, compared with one point one billion euros in 2025.
22:48
Growth and restructuring can happen at the same time.
22:54
Sereact announced in July that Zalando had joined its ongoing Series B,
23:02
increasing the round total from the previously announced one hundred and ten
23:10
million dollars to one hundred and sixteen million dollars.
23:16
The accurate July increment is six million dollars, not a fresh one
23:24
hundred and sixteen million dollar round.
23:28
QuantumDiamonds announced a ninety-one million euro financing package in July.
23:34
Fifteen million euros was equity led by World Fund; seventy-six million euros
23:42
was approved non-dilutive public funding under the European Chips Act.
23:49
It is a significant package, but only the equity component belongs in
23:57
a venture-capital total.
23:59
And the Startup Association's formation data prevents us from declaring
24:05
software dead. Of the three thousand and fifty-three startups founded in the
24:13
first half, eight hundred and forty-four were software companies.
24:19
Thirty-four percent of all new startups had a clear AI connection.
24:26
Germany is adding industrial capacity without ceasing to create software
24:33
companies.
24:34
Three takeaways.
24:35
First, founders should distinguish between a sector-wide funding environment
24:40
and a handful of mega-rounds. July shows exceptional demand for selected
24:47
defence, autonomy, and fusion companies. It does not show that every hard-tech
24:54
startup can raise easily or that software capital has disappeared.
25:00
Second, investors should separate announced exit value from realised
25:06
liquidity. Delivery Hero and AtaiBeckley can become important distributions
25:11
only if the transactions close. Prior Labs is the completed acquisition, but
25:19
its purchase price remains undisclosed.
25:22
Third, everyone analysing the ecosystem should keep the capital buckets
25:28
separate. Startup equity, secondary transactions, public grants, and fund
25:33
closures answer different questions. Adding them together creates a larger
25:39
headline and a weaker analysis.
25:43
Watch four things.
25:46
One: whether Isar Aerospace completes a successful orbital flight before
25:58
year-end.
25:59
Two: whether and when KNDS revives its IPO plan after the July
26:13
postponement.
26:14
Three: the acceptance and regulatory process for Uber's Delivery Hero offer,
26:27
which is not expected to close before the second half of 2027.
26:41
Four: shareholder and regulatory progress on Lilly's proposed AtaiBeckley
26:52
acquisition, including whether any contingent-value milestones eventually
27:00
become payable.
27:03
I am keeping two predictions on record.
27:09
Prediction one: Helsing reaches a twenty-five billion dollar valuation within
27:19
twelve months. That is a forecast, not a fact, and it depends
27:31
on procurement growth, execution, and investor appetite remaining strong.
27:40
Prediction two: at least two more German defence startups reach unicorn status
27:51
before the end of 2026. Again, that is my forecast, not a
28:03
reported outcome.
28:05
The July evidence supports a narrower and stronger conclusion than the
28:16
original headline. Germany has not stopped building software.
28:24
But its largest July cheques went to defence AI, autonomous systems, and
28:35
fusion. That is where the capital rotation is most visible.
28:45
If you found this useful, share it with one person who works
28:57
in European venture, deep tech, or defence technology.